-The figure here displays a simple example of an RNN, with inputs \( x_t \)
-at a given time \( t \) and outputs \( y_t \). Introducing time as a variable
-offers an intutitive way of understanding these networks. In addition
-to the inputs \( x_t \), the layer at a time \( t \) receives also as input
-the output from the previous layer \( t-1 \), that is \( y_{t1} \).
+The following code provides an example of how recurrent neural
+networks can be used to extrapolate to unknown values of physics data
+sets. Specifically, the data sets used in this program come from
+a quantum mechanical many-body calculation of energies as functions of the number of particles.
-This means also that we need to have weights that link both the inputs
-\( x_t \) to the outputs \( y_t \) as well as weights that link the output
-from the previous time \( y_{t-1} \) and \( y_t \). The figure here shows an
-example of a simple RNN.
-
-More material will be added here.
+
+
# For matrices and calculations
+importnumpyasnp
+# For machine learning (backend for keras)
+importtensorflowastf
+# User-friendly machine learning library
+# Front end for TensorFlow
+importtensorflow.keras
+# Different methods from Keras needed to create an RNN
+# This is not necessary but it shortened function calls
+# that need to be used in the code.
+fromtensorflow.kerasimport datasets, layers, models
+fromtensorflow.keras.layersimport Input
+fromtensorflow.kerasimport regularizers
+fromtensorflow.keras.modelsimport Model, Sequential
+fromtensorflow.keras.layersimport Dense, SimpleRNN, LSTM, GRU
+# For timing the code
+fromtimeitimport default_timer as timer
+# For plotting
+importmatplotlib.pyplotasplt
+
+
+# The data set
+datatype='VaryDimension'
+X_tot = np.arange(2, 42, 2)
+y_tot = np.array([-0.03077640549, -0.08336233266, -0.1446729567, -0.2116753732, -0.2830637392, -0.3581341341, -0.436462435, -0.5177783846,
+ -0.6019067271, -0.6887363571, -0.7782028952, -0.8702784034, -0.9649652536, -1.062292565, -1.16231451,
+ -1.265109911, -1.370782966, -1.479465113, -1.591317992, -1.70653767])
+
-
Solving differential equations and eigenvalue problems with RNNs
+
Formatting the Data
-In our discussions of ordinary differential equations and partial
-differential equations using neural networks. Here we will discuss how
-we can solve say ordinary differential equations and eigenvalue
-problems using RNNs. Eigenvalue problems can be solved using RNNs by
-rewriting such a problems as a non-linear differential equation.
+The way the recurrent neural networks are trained in this program
+differs from how machine learning algorithms are usually trained.
+Typically a machine learning algorithm is trained by learning the
+relationship between the x data and the y data. In this program, the
+recurrent neural network will be trained to recognize the relationship
+in a sequence of y values. This is type of data formatting is
+typically used time series forcasting, but it can also be used in any
+extrapolation (time series forecasting is just a specific type of
+extrapolation along the time axis). This method of data formatting
+does not use the x data and assumes that the y data are evenly spaced.
-Instead of starting with a well-known ordinary differential equation,
-we start directly with an eigenvaule problem.
+For a standard machine learning algorithm, the training data has the
+form of (x,y) so the machine learning algorithm learns to assiciate a
+y value with a given x value. This is useful when the test data has x
+values within the same range as the training data. However, for this
+application, the x values of the test data are outside of the x values
+of the training data and the traditional method of training a machine
+learning algorithm does not work as well. For this reason, the
+recurrent neural network is trained on sequences of y values of the
+form ((y1, y2), y3), so that the network is concerned with learning
+the pattern of the y data and not the relation between the x and y
+data. As long as the pattern of y data outside of the training region
+stays relatively stable compared to what was inside the training
+region, this method of training can produce accurate extrapolations to
+y values far removed from the training data set.
+
+
+
+
+
+
+
+
+
+
+
+
+
# FORMAT_DATA
+defformat_data(data, length_of_sequence = 2):
+ """
+ Inputs:
+ data(a numpy array): the data that will be the inputs to the recurrent neural
+ network
+ length_of_sequence (an int): the number of elements in one iteration of the
+ sequence patter. For a function approximator use length_of_sequence = 2.
+ Returns:
+ rnn_input (a 3D numpy array): the input data for the recurrent neural network. Its
+ dimensions are length of data - length of sequence, length of sequence,
+ dimnsion of data
+ rnn_output (a numpy array): the training data for the neural network
+ Formats data to be used in a recurrent neural network.
+ """
+
+ X, Y = [], []
+ for i inrange(len(data)-length_of_sequence):
+ # Get the next length_of_sequence elements
+ a = data[i:i+length_of_sequence]
+ # Get the element that immediately follows that
+ b = data[i+length_of_sequence]
+ # Reshape so that each data point is contained in its own array
+ a = np.reshape (a, (len(a), 1))
+ X.append(a)
+ Y.append(b)
+ rnn_input = np.array(X)
+ rnn_output = np.array(Y)
+
+ return rnn_input, rnn_output
+
+
+# ## Defining the Recurrent Neural Network Using Keras
+#
+# The following method defines a simple recurrent neural network in keras consisting of one input layer, one hidden layer, and one output layer.
+
+defrnn(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with one hidden layer and returns the model.
+ """
+ # Number of neurons in the input and output layers
+ in_out_neurons = 1
+ # Number of neurons in the hidden layer
+ hidden_neurons = 200
+ # Define the input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Define the hidden layer as a simple RNN layer with a set number of neurons and add it to
+ # the network immediately after the input layer
+ rnn = SimpleRNN(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN")(inp)
+ # Define the output layer as a dense neural network layer (standard neural network layer)
+ #and add it to the network immediately after the hidden layer.
+ dens = Dense(in_out_neurons,name="dense")(rnn)
+ # Create the machine learning model starting with the input layer and ending with the
+ # output layer
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the machine learning model using the mean squared error function as the loss
+ # function and an Adams optimizer.
+ model.compile(loss="mean_squared_error", optimizer="adam")
+ return model
+
-
Long-Short Time Memory
+
Predicting New Points With A Trained Recurrent Neural Network
-Discussions about dynamic unrolling through time. discuss memory cells, input and output
+
+
+
deftest_rnn (x1, y_test, plot_min, plot_max):
+ """
+ Inputs:
+ x1 (a list or numpy array): The complete x component of the data set
+ y_test (a list or numpy array): The complete y component of the data set
+ plot_min (an int or float): the smallest x value used in the training data
+ plot_max (an int or float): the largest x valye used in the training data
+ Returns:
+ None.
+ Uses a trained recurrent neural network model to predict future points in the
+ series. Computes the MSE of the predicted data set from the true data set, saves
+ the predicted data set to a csv file, and plots the predicted and true data sets w
+ while also displaying the data range used for training.
+ """
+ # Add the training data as the first dim points in the predicted data array as these
+ # are known values.
+ y_pred = y_test[:dim].tolist()
+ # Generate the first input to the trained recurrent neural network using the last two
+ # points of the training data. Based on how the network was trained this means that it
+ # will predict the first point in the data set after the training data. All of the
+ # brackets are necessary for Tensorflow.
+ next_input = np.array([[[y_test[dim-2]], [y_test[dim-1]]]])
+ # Save the very last point in the training data set. This will be used later.
+ last = [y_test[dim-1]]
+
+ # Iterate until the complete data set is created.
+ for i inrange (dim, len(y_test)):
+ # Predict the next point in the data set using the previous two points.
+ next = model.predict(next_input)
+ # Append just the number of the predicted data set
+ y_pred.append(next[0][0])
+ # Create the input that will be used to predict the next data point in the data set.
+ next_input = np.array([[last, next[0]]], dtype=np.float64)
+ last = next
+
+ # Print the mean squared error between the known data set and the predicted data set.
+ print('MSE: ', np.square(np.subtract(y_test, y_pred)).mean())
+ # Save the predicted data set as a csv file for later use
+ name = datatype + 'Predicted'+str(dim)+'.csv'
+ np.savetxt(name, y_pred, delimiter=',')
+ # Plot the known data set and the predicted data set. The red box represents the region that was used
+ # for the training data.
+ fig, ax = plt.subplots()
+ ax.plot(x1, y_test, label="true", linewidth=3)
+ ax.plot(x1, y_pred, 'g-.',label="predicted", linewidth=4)
+ ax.legend()
+ # Created a red region to represent the points used in the training data.
+ ax.axvspan(plot_min, plot_max, alpha=0.25, color='red')
+ plt.show()
+
+# Check to make sure the data set is complete
+assertlen(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+model = rnn(length_of_sequences = rnn_input.shape[1])
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
-
Autoencoders: Overarching view
+
Other Things to Try
-Autoencoders are artificial neural networks capable of learning
-efficient representations of the input data (these representations are called codings) without
-any supervision (i.e., the training set is unlabeled). These codings
-typically have a much lower dimensionality than the input data, making
-autoencoders useful for dimensionality reduction.
+Changing the size of the recurrent neural network and its parameters
+can drastically change the results you get from the model. The below
+code takes the simple recurrent neural network from above and adds a
+second hidden layer, changes the number of neurons in the hidden
+layer, and explicitly declares the activation function of the hidden
+layers to be a sigmoid function. The loss function and optimizer can
+also be changed but are kept the same as the above network. These
+parameters can be tuned to provide the optimal result from the
+network. For some ideas on how to improve the performance of a
+recurrent neural network.
-More importantly, autoencoders act as powerful feature detectors, and
-they can be used for unsupervised pretraining of deep neural networks.
-
-Lastly, they are capable of randomly generating new data that looks
-very similar to the training data; this is called a generative
-model. For example, you could train an autoencoder on pictures of
-faces, and it would then be able to generate new faces. Surprisingly,
-autoencoders work by simply learning to copy their inputs to their
-outputs. This may sound like a trivial task, but we will see that
-constraining the network in various ways can make it rather
-difficult. For example, you can limit the size of the internal
-representation, or you can add noise to the inputs and train the
-network to recover the original inputs. These constraints prevent the
-autoencoder from trivially copying the inputs directly to the outputs,
-which forces it to learn efficient ways of representing the data. In
-short, the codings are byproducts of the autoencoder’s attempt to
-learn the identity function under some constraints.
+
+
defrnn_2layers(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with two hidden layers and returns the model.
+ """
+ # Number of neurons in the input and output layers
+ in_out_neurons = 1
+ # Number of neurons in the hidden layer, increased from the first network
+ hidden_neurons = 500
+ # Define the input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Create two hidden layers instead of one hidden layer. Explicitly set the activation
+ # function to be the sigmoid function (the default value is hyperbolic tangent)
+ rnn1 = SimpleRNN(hidden_neurons,
+ return_sequences=True, # This needs to be True if another hidden layer is to follow
+ stateful = stateful, activation = 'sigmoid',
+ name="RNN1")(inp)
+ rnn2 = SimpleRNN(hidden_neurons,
+ return_sequences=False, activation = 'sigmoid',
+ stateful = stateful,
+ name="RNN2")(rnn1)
+ # Define the output layer as a dense neural network layer (standard neural network layer)
+ #and add it to the network immediately after the hidden layer.
+ dens = Dense(in_out_neurons,name="dense")(rnn2)
+ # Create the machine learning model starting with the input layer and ending with the
+ # output layer
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the machine learning model using the mean squared error function as the loss
+ # function and an Adams optimizer.
+ model.compile(loss="mean_squared_error", optimizer="adam")
+ return model
+
+# Check to make sure the data set is complete
+assertlen(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+model = rnn_2layers(length_of_sequences = 2)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
+The first network created below is similar to the previous network,
+but it replaces the SimpleRNN layers with LSTM layers. The second
+network below has two hidden layers made up of GRUs, which are
+preceeded by two dense (feeddorward) neural network layers. These
+dense layers "preprocess" the data before it reaches the recurrent
+layers. This architecture has been shown to improve the performance
+of recurrent neural networks (see the link above and also
+https://arxiv.org/pdf/1807.02857.pdf.
+
+
+
+
+
deflstm_2layers(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with two LSTM hidden layers and returns the model.
+ """
+ # Number of neurons on the input/output layer and the number of neurons in the hidden layer
+ in_out_neurons = 1
+ hidden_neurons = 250
+ # Input Layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Hidden layers (in this case they are LSTM layers instead if SimpleRNN layers)
+ rnn= LSTM(hidden_neurons,
+ return_sequences=True,
+ stateful = stateful,
+ name="RNN", use_bias=True, activation='tanh')(inp)
+ rnn1 = LSTM(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN1", use_bias=True, activation='tanh')(rnn)
+ # Output layer
+ dens = Dense(in_out_neurons,name="dense")(rnn1)
+ # Define the midel
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the model
+ model.compile(loss='mean_squared_error', optimizer='adam')
+ # Return the model
+ return model
+
+defdnn2_gru2(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with four hidden layers (two dense followed by
+ two GRU layers) and returns the model.
+ """
+ # Number of neurons on the input/output layers and hidden layers
+ in_out_neurons = 1
+ hidden_neurons = 250
+ # Input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Hidden Dense (feedforward) layers
+ dnn = Dense(hidden_neurons/2, activation='relu', name='dnn')(inp)
+ dnn1 = Dense(hidden_neurons/2, activation='relu', name='dnn1')(dnn)
+ # Hidden GRU layers
+ rnn1 = GRU(hidden_neurons,
+ return_sequences=True,
+ stateful = stateful,
+ name="RNN1", use_bias=True)(dnn1)
+ rnn = GRU(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN", use_bias=True)(rnn1)
+ # Output layer
+ dens = Dense(in_out_neurons,name="dense")(rnn)
+ # Define the model
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the mdoel
+ model.compile(loss='mean_squared_error', optimizer='adam')
+ # Return the model
+ return model
+
+# Check to make sure the data set is complete
+assertlen(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+# Change the method name to reflect which network you want to use
+model = dnn2_gru2(length_of_sequences = 2)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
+
+# ### Training Recurrent Neural Networks in the Standard Way (i.e. learning the relationship between the X and Y data)
+#
+# Finally, comparing the performace of a recurrent neural network using the standard data formatting to the performance of the network with time sequence data formatting shows the benefit of this type of data formatting with extrapolation.
+
+# Check to make sure the data set is complete
+assertlen(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+# Reshape the data for Keras specifications
+X_train = X_train.reshape((dim, 1))
+y_train = y_train.reshape((dim, 1))
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+# Set the sequence length to 1 for regular data formatting
+model = rnn(length_of_sequences = 1)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(X_train, y_train, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict the remaining data points
+X_pred = X_tot[dim:]
+X_pred = X_pred.reshape((len(X_pred), 1))
+y_model = model.predict(X_pred)
+y_pred = np.concatenate((y_tot[:dim], y_model.flatten()))
+
+# Plot the known data set and the predicted data set. The red box represents the region that was used
+# for the training data.
+fig, ax = plt.subplots()
+ax.plot(X_tot, y_tot, label="true", linewidth=3)
+ax.plot(X_tot, y_pred, 'g-.',label="predicted", linewidth=4)
+ax.legend()
+# Created a red region to represent the points used in the training data.
+ax.axvspan(X_tot[0], X_tot[dim], alpha=0.25, color='red')
+plt.show()
+
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
diff --git a/doc/pub/week42/html/week42-solarized.html b/doc/pub/week42/html/week42-solarized.html
index a7c80e788..d00badbb0 100644
--- a/doc/pub/week42/html/week42-solarized.html
+++ b/doc/pub/week42/html/week42-solarized.html
@@ -7,9 +7,9 @@ Automatically generated HTML file from DocOnce source
-
+
-
Week 42 Convolutional and Recurrent Neural Networks and Autoencoders
+Week 42 Convolutional (CNN) and Recurrent (RNN) Neural Networks
@@ -26,6 +26,32 @@ pre {
border: 0pt solid #93a1a1;
box-shadow: none;
}
+.alert-text-small { font-size: 80%; }
+.alert-text-large { font-size: 130%; }
+.alert-text-normal { font-size: 90%; }
+.alert {
+ padding:8px 35px 8px 14px; margin-bottom:18px;
+ text-shadow:0 1px 0 rgba(255,255,255,0.5);
+ border:1px solid #93a1a1;
+ border-radius: 4px;
+ -webkit-border-radius: 4px;
+ -moz-border-radius: 4px;
+ color: #555;
+ background-color: #eee8d5;
+ background-position: 10px 5px;
+ background-repeat: no-repeat;
+ background-size: 38px;
+ padding-left: 55px;
+ width: 75%;
+ }
+.alert-block {padding-top:14px; padding-bottom:14px}
+.alert-block > p, .alert-block > ul {margin-bottom:1em}
+.alert li {margin-top: 1em}
+.alert-block p+p {margin-top:5px}
+.alert-notice { background-image: url(https://cdn.rawgit.com/hplgit/doconce/master/bundled/html_images/small_yellow_notice.png); }
+.alert-summary { background-image:url(https://cdn.rawgit.com/hplgit/doconce/master/bundled/html_images/small_yellow_summary.png); }
+.alert-warning { background-image: url(https://cdn.rawgit.com/hplgit/doconce/master/bundled/html_images/small_yellow_warning.png); }
+.alert-question {background-image:url(https://cdn.rawgit.com/hplgit/doconce/master/bundled/html_images/small_yellow_question.png); }
div { text-align: justify; text-justify: inter-word; }
@@ -82,16 +108,19 @@ div { text-align: justify; text-justify: inter-word; }
2,
None,
'___sec26'),
- ('A simple example', 2, None, '___sec27'),
- ('Set up of an RNN', 2, None, '___sec28'),
- ('Solving differential equations and eigenvalue problems with '
- 'RNNs',
+ ('Set up of an RNN', 2, None, '___sec27'),
+ ('A simple example', 2, None, '___sec28'),
+ ('An extrapolation example', 2, None, '___sec29'),
+ ('Formatting the Data', 2, None, '___sec30'),
+ ('Predicting New Points With A Trained Recurrent Neural Network',
2,
None,
- '___sec29'),
- ('Long-Short Time Memory', 2, None, '___sec30'),
- ('Autoencoders: Overarching view', 2, None, '___sec31'),
- ('Simple examples of Autoencoders', 2, None, '___sec32')]}
+ '___sec31'),
+ ('Other Things to Try', 2, None, '___sec32'),
+ ('Other Types of Recurrent Neural Networks',
+ 2,
+ None,
+ '___sec33')]}
end of tocinfo -->
@@ -117,7 +146,7 @@ MathJax.Hub.Config({
-
Week 42 Convolutional and Recurrent Neural Networks and Autoencoders
+
Week 42 Convolutional (CNN) and Recurrent (RNN) Neural Networks
@@ -133,7 +162,7 @@ MathJax.Hub.Config({
[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University
-
Oct 15, 2020
+
Oct 16, 2020
@@ -141,12 +170,24 @@ MathJax.Hub.Config({
Plan for week 42
-
Thursday: Convolutional Neural Networks and examples
-
Friday: Recurrent Neural Networks and Autoencoders
+
Thursday: Convolutional Neural Networks and examples. Video of Lecture
-The figure here displays a simple example of an RNN, with inputs \( x_t \)
-at a given time \( t \) and outputs \( y_t \). Introducing time as a variable
-offers an intutitive way of understanding these networks. In addition
-to the inputs \( x_t \), the layer at a time \( t \) receives also as input
-the output from the previous layer \( t-1 \), that is \( y_{t1} \).
+The following code provides an example of how recurrent neural
+networks can be used to extrapolate to unknown values of physics data
+sets. Specifically, the data sets used in this program come from
+a quantum mechanical many-body calculation of energies as functions of the number of particles.
-This means also that we need to have weights that link both the inputs
-\( x_t \) to the outputs \( y_t \) as well as weights that link the output
-from the previous time \( y_{t-1} \) and \( y_t \). The figure here shows an
-example of a simple RNN.
-
-More material will be added here.
+
+
# For matrices and calculations
+importnumpyasnp
+# For machine learning (backend for keras)
+importtensorflowastf
+# User-friendly machine learning library
+# Front end for TensorFlow
+importtensorflow.keras
+# Different methods from Keras needed to create an RNN
+# This is not necessary but it shortened function calls
+# that need to be used in the code.
+fromtensorflow.kerasimport datasets, layers, models
+fromtensorflow.keras.layersimport Input
+fromtensorflow.kerasimport regularizers
+fromtensorflow.keras.modelsimport Model, Sequential
+fromtensorflow.keras.layersimport Dense, SimpleRNN, LSTM, GRU
+# For timing the code
+fromtimeitimport default_timer as timer
+# For plotting
+importmatplotlib.pyplotasplt
+
+# The data set
+datatype='VaryDimension'
+X_tot = np.arange(2, 42, 2)
+y_tot = np.array([-0.03077640549, -0.08336233266, -0.1446729567, -0.2116753732, -0.2830637392, -0.3581341341, -0.436462435, -0.5177783846,
+ -0.6019067271, -0.6887363571, -0.7782028952, -0.8702784034, -0.9649652536, -1.062292565, -1.16231451,
+ -1.265109911, -1.370782966, -1.479465113, -1.591317992, -1.70653767])
+
-
Solving differential equations and eigenvalue problems with RNNs
+
Formatting the Data
-In our discussions of ordinary differential equations and partial
-differential equations using neural networks. Here we will discuss how
-we can solve say ordinary differential equations and eigenvalue
-problems using RNNs. Eigenvalue problems can be solved using RNNs by
-rewriting such a problems as a non-linear differential equation.
+The way the recurrent neural networks are trained in this program
+differs from how machine learning algorithms are usually trained.
+Typically a machine learning algorithm is trained by learning the
+relationship between the x data and the y data. In this program, the
+recurrent neural network will be trained to recognize the relationship
+in a sequence of y values. This is type of data formatting is
+typically used time series forcasting, but it can also be used in any
+extrapolation (time series forecasting is just a specific type of
+extrapolation along the time axis). This method of data formatting
+does not use the x data and assumes that the y data are evenly spaced.
-Instead of starting with a well-known ordinary differential equation,
-we start directly with an eigenvaule problem.
+For a standard machine learning algorithm, the training data has the
+form of (x,y) so the machine learning algorithm learns to assiciate a
+y value with a given x value. This is useful when the test data has x
+values within the same range as the training data. However, for this
+application, the x values of the test data are outside of the x values
+of the training data and the traditional method of training a machine
+learning algorithm does not work as well. For this reason, the
+recurrent neural network is trained on sequences of y values of the
+form ((y1, y2), y3), so that the network is concerned with learning
+the pattern of the y data and not the relation between the x and y
+data. As long as the pattern of y data outside of the training region
+stays relatively stable compared to what was inside the training
+region, this method of training can produce accurate extrapolations to
+y values far removed from the training data set.
+
+
+
+
+
+
+
+
+
+
+
+
# FORMAT_DATA
+defformat_data(data, length_of_sequence = 2):
+ """
+ Inputs:
+ data(a numpy array): the data that will be the inputs to the recurrent neural
+ network
+ length_of_sequence (an int): the number of elements in one iteration of the
+ sequence patter. For a function approximator use length_of_sequence = 2.
+ Returns:
+ rnn_input (a 3D numpy array): the input data for the recurrent neural network. Its
+ dimensions are length of data - length of sequence, length of sequence,
+ dimnsion of data
+ rnn_output (a numpy array): the training data for the neural network
+ Formats data to be used in a recurrent neural network.
+ """
+
+ X, Y = [], []
+ for i inrange(len(data)-length_of_sequence):
+ # Get the next length_of_sequence elements
+ a = data[i:i+length_of_sequence]
+ # Get the element that immediately follows that
+ b = data[i+length_of_sequence]
+ # Reshape so that each data point is contained in its own array
+ a = np.reshape (a, (len(a), 1))
+ X.append(a)
+ Y.append(b)
+ rnn_input = np.array(X)
+ rnn_output = np.array(Y)
+
+ return rnn_input, rnn_output
+
+
+# ## Defining the Recurrent Neural Network Using Keras
+#
+# The following method defines a simple recurrent neural network in keras consisting of one input layer, one hidden layer, and one output layer.
+
+defrnn(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with one hidden layer and returns the model.
+ """
+ # Number of neurons in the input and output layers
+ in_out_neurons = 1
+ # Number of neurons in the hidden layer
+ hidden_neurons = 200
+ # Define the input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Define the hidden layer as a simple RNN layer with a set number of neurons and add it to
+ # the network immediately after the input layer
+ rnn = SimpleRNN(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN")(inp)
+ # Define the output layer as a dense neural network layer (standard neural network layer)
+ #and add it to the network immediately after the hidden layer.
+ dens = Dense(in_out_neurons,name="dense")(rnn)
+ # Create the machine learning model starting with the input layer and ending with the
+ # output layer
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the machine learning model using the mean squared error function as the loss
+ # function and an Adams optimizer.
+ model.compile(loss="mean_squared_error", optimizer="adam")
+ return model
+
-
Long-Short Time Memory
+
Predicting New Points With A Trained Recurrent Neural Network
-Discussions about dynamic unrolling through time. discuss memory cells, input and output
+
+
deftest_rnn (x1, y_test, plot_min, plot_max):
+ """
+ Inputs:
+ x1 (a list or numpy array): The complete x component of the data set
+ y_test (a list or numpy array): The complete y component of the data set
+ plot_min (an int or float): the smallest x value used in the training data
+ plot_max (an int or float): the largest x valye used in the training data
+ Returns:
+ None.
+ Uses a trained recurrent neural network model to predict future points in the
+ series. Computes the MSE of the predicted data set from the true data set, saves
+ the predicted data set to a csv file, and plots the predicted and true data sets w
+ while also displaying the data range used for training.
+ """
+ # Add the training data as the first dim points in the predicted data array as these
+ # are known values.
+ y_pred = y_test[:dim].tolist()
+ # Generate the first input to the trained recurrent neural network using the last two
+ # points of the training data. Based on how the network was trained this means that it
+ # will predict the first point in the data set after the training data. All of the
+ # brackets are necessary for Tensorflow.
+ next_input = np.array([[[y_test[dim-2]], [y_test[dim-1]]]])
+ # Save the very last point in the training data set. This will be used later.
+ last = [y_test[dim-1]]
+
+ # Iterate until the complete data set is created.
+ for i inrange (dim, len(y_test)):
+ # Predict the next point in the data set using the previous two points.
+ next = model.predict(next_input)
+ # Append just the number of the predicted data set
+ y_pred.append(next[0][0])
+ # Create the input that will be used to predict the next data point in the data set.
+ next_input = np.array([[last, next[0]]], dtype=np.float64)
+ last = next
+
+ # Print the mean squared error between the known data set and the predicted data set.
+ print('MSE: ', np.square(np.subtract(y_test, y_pred)).mean())
+ # Save the predicted data set as a csv file for later use
+ name = datatype + 'Predicted'+str(dim)+'.csv'
+ np.savetxt(name, y_pred, delimiter=',')
+ # Plot the known data set and the predicted data set. The red box represents the region that was used
+ # for the training data.
+ fig, ax = plt.subplots()
+ ax.plot(x1, y_test, label="true", linewidth=3)
+ ax.plot(x1, y_pred, 'g-.',label="predicted", linewidth=4)
+ ax.legend()
+ # Created a red region to represent the points used in the training data.
+ ax.axvspan(plot_min, plot_max, alpha=0.25, color='red')
+ plt.show()
+
+# Check to make sure the data set is complete
+assertlen(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+model = rnn(length_of_sequences = rnn_input.shape[1])
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
-
Autoencoders: Overarching view
+
Other Things to Try
-Autoencoders are artificial neural networks capable of learning
-efficient representations of the input data (these representations are called codings) without
-any supervision (i.e., the training set is unlabeled). These codings
-typically have a much lower dimensionality than the input data, making
-autoencoders useful for dimensionality reduction.
+Changing the size of the recurrent neural network and its parameters
+can drastically change the results you get from the model. The below
+code takes the simple recurrent neural network from above and adds a
+second hidden layer, changes the number of neurons in the hidden
+layer, and explicitly declares the activation function of the hidden
+layers to be a sigmoid function. The loss function and optimizer can
+also be changed but are kept the same as the above network. These
+parameters can be tuned to provide the optimal result from the
+network. For some ideas on how to improve the performance of a
+recurrent neural network.
-More importantly, autoencoders act as powerful feature detectors, and
-they can be used for unsupervised pretraining of deep neural networks.
-
-Lastly, they are capable of randomly generating new data that looks
-very similar to the training data; this is called a generative
-model. For example, you could train an autoencoder on pictures of
-faces, and it would then be able to generate new faces. Surprisingly,
-autoencoders work by simply learning to copy their inputs to their
-outputs. This may sound like a trivial task, but we will see that
-constraining the network in various ways can make it rather
-difficult. For example, you can limit the size of the internal
-representation, or you can add noise to the inputs and train the
-network to recover the original inputs. These constraints prevent the
-autoencoder from trivially copying the inputs directly to the outputs,
-which forces it to learn efficient ways of representing the data. In
-short, the codings are byproducts of the autoencoder’s attempt to
-learn the identity function under some constraints.
+
+
defrnn_2layers(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with two hidden layers and returns the model.
+ """
+ # Number of neurons in the input and output layers
+ in_out_neurons = 1
+ # Number of neurons in the hidden layer, increased from the first network
+ hidden_neurons = 500
+ # Define the input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Create two hidden layers instead of one hidden layer. Explicitly set the activation
+ # function to be the sigmoid function (the default value is hyperbolic tangent)
+ rnn1 = SimpleRNN(hidden_neurons,
+ return_sequences=True, # This needs to be True if another hidden layer is to follow
+ stateful = stateful, activation = 'sigmoid',
+ name="RNN1")(inp)
+ rnn2 = SimpleRNN(hidden_neurons,
+ return_sequences=False, activation = 'sigmoid',
+ stateful = stateful,
+ name="RNN2")(rnn1)
+ # Define the output layer as a dense neural network layer (standard neural network layer)
+ #and add it to the network immediately after the hidden layer.
+ dens = Dense(in_out_neurons,name="dense")(rnn2)
+ # Create the machine learning model starting with the input layer and ending with the
+ # output layer
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the machine learning model using the mean squared error function as the loss
+ # function and an Adams optimizer.
+ model.compile(loss="mean_squared_error", optimizer="adam")
+ return model
+# Check to make sure the data set is complete
+assertlen(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+model = rnn_2layers(length_of_sequences = 2)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
+The first network created below is similar to the previous network,
+but it replaces the SimpleRNN layers with LSTM layers. The second
+network below has two hidden layers made up of GRUs, which are
+preceeded by two dense (feeddorward) neural network layers. These
+dense layers "preprocess" the data before it reaches the recurrent
+layers. This architecture has been shown to improve the performance
+of recurrent neural networks (see the link above and also
+https://arxiv.org/pdf/1807.02857.pdf.
+
+
+
+
+
deflstm_2layers(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with two LSTM hidden layers and returns the model.
+ """
+ # Number of neurons on the input/output layer and the number of neurons in the hidden layer
+ in_out_neurons = 1
+ hidden_neurons = 250
+ # Input Layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Hidden layers (in this case they are LSTM layers instead if SimpleRNN layers)
+ rnn= LSTM(hidden_neurons,
+ return_sequences=True,
+ stateful = stateful,
+ name="RNN", use_bias=True, activation='tanh')(inp)
+ rnn1 = LSTM(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN1", use_bias=True, activation='tanh')(rnn)
+ # Output layer
+ dens = Dense(in_out_neurons,name="dense")(rnn1)
+ # Define the midel
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the model
+ model.compile(loss='mean_squared_error', optimizer='adam')
+ # Return the model
+ return model
+
+defdnn2_gru2(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with four hidden layers (two dense followed by
+ two GRU layers) and returns the model.
+ """
+ # Number of neurons on the input/output layers and hidden layers
+ in_out_neurons = 1
+ hidden_neurons = 250
+ # Input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Hidden Dense (feedforward) layers
+ dnn = Dense(hidden_neurons/2, activation='relu', name='dnn')(inp)
+ dnn1 = Dense(hidden_neurons/2, activation='relu', name='dnn1')(dnn)
+ # Hidden GRU layers
+ rnn1 = GRU(hidden_neurons,
+ return_sequences=True,
+ stateful = stateful,
+ name="RNN1", use_bias=True)(dnn1)
+ rnn = GRU(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN", use_bias=True)(rnn1)
+ # Output layer
+ dens = Dense(in_out_neurons,name="dense")(rnn)
+ # Define the model
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the mdoel
+ model.compile(loss='mean_squared_error', optimizer='adam')
+ # Return the model
+ return model
+
+# Check to make sure the data set is complete
+assertlen(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+# Change the method name to reflect which network you want to use
+model = dnn2_gru2(length_of_sequences = 2)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
+
+# ### Training Recurrent Neural Networks in the Standard Way (i.e. learning the relationship between the X and Y data)
+#
+# Finally, comparing the performace of a recurrent neural network using the standard data formatting to the performance of the network with time sequence data formatting shows the benefit of this type of data formatting with extrapolation.
+
+# Check to make sure the data set is complete
+assertlen(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+# Reshape the data for Keras specifications
+X_train = X_train.reshape((dim, 1))
+y_train = y_train.reshape((dim, 1))
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+# Set the sequence length to 1 for regular data formatting
+model = rnn(length_of_sequences = 1)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(X_train, y_train, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict the remaining data points
+X_pred = X_tot[dim:]
+X_pred = X_pred.reshape((len(X_pred), 1))
+y_model = model.predict(X_pred)
+y_pred = np.concatenate((y_tot[:dim], y_model.flatten()))
+
+# Plot the known data set and the predicted data set. The red box represents the region that was used
+# for the training data.
+fig, ax = plt.subplots()
+ax.plot(X_tot, y_tot, label="true", linewidth=3)
+ax.plot(X_tot, y_pred, 'g-.',label="predicted", linewidth=4)
+ax.legend()
+# Created a red region to represent the points used in the training data.
+ax.axvspan(X_tot[0], X_tot[dim], alpha=0.25, color='red')
+plt.show()
+
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
diff --git a/doc/pub/week42/html/week42.html b/doc/pub/week42/html/week42.html
index 29eca21d4..8c8c688c6 100644
--- a/doc/pub/week42/html/week42.html
+++ b/doc/pub/week42/html/week42.html
@@ -7,9 +7,9 @@ Automatically generated HTML file from DocOnce source
-
+
-
Week 42 Convolutional and Recurrent Neural Networks and Autoencoders
+Week 42 Convolutional (CNN) and Recurrent (RNN) Neural Networks
@@ -87,16 +113,19 @@ div { text-align: justify; text-justify: inter-word; }
2,
None,
'___sec26'),
- ('A simple example', 2, None, '___sec27'),
- ('Set up of an RNN', 2, None, '___sec28'),
- ('Solving differential equations and eigenvalue problems with '
- 'RNNs',
+ ('Set up of an RNN', 2, None, '___sec27'),
+ ('A simple example', 2, None, '___sec28'),
+ ('An extrapolation example', 2, None, '___sec29'),
+ ('Formatting the Data', 2, None, '___sec30'),
+ ('Predicting New Points With A Trained Recurrent Neural Network',
2,
None,
- '___sec29'),
- ('Long-Short Time Memory', 2, None, '___sec30'),
- ('Autoencoders: Overarching view', 2, None, '___sec31'),
- ('Simple examples of Autoencoders', 2, None, '___sec32')]}
+ '___sec31'),
+ ('Other Things to Try', 2, None, '___sec32'),
+ ('Other Types of Recurrent Neural Networks',
+ 2,
+ None,
+ '___sec33')]}
end of tocinfo -->
@@ -122,7 +151,7 @@ MathJax.Hub.Config({
-
Week 42 Convolutional and Recurrent Neural Networks and Autoencoders
+
Week 42 Convolutional (CNN) and Recurrent (RNN) Neural Networks
@@ -138,7 +167,7 @@ MathJax.Hub.Config({
[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University
-
Oct 15, 2020
+
Oct 16, 2020
@@ -146,12 +175,24 @@ MathJax.Hub.Config({
Plan for week 42
-
Thursday: Convolutional Neural Networks and examples
-
Friday: Recurrent Neural Networks and Autoencoders
+
Thursday: Convolutional Neural Networks and examples. Video of Lecture
-The figure here displays a simple example of an RNN, with inputs \( x_t \)
-at a given time \( t \) and outputs \( y_t \). Introducing time as a variable
-offers an intutitive way of understanding these networks. In addition
-to the inputs \( x_t \), the layer at a time \( t \) receives also as input
-the output from the previous layer \( t-1 \), that is \( y_{t1} \).
+The following code provides an example of how recurrent neural
+networks can be used to extrapolate to unknown values of physics data
+sets. Specifically, the data sets used in this program come from
+a quantum mechanical many-body calculation of energies as functions of the number of particles.
-This means also that we need to have weights that link both the inputs
-\( x_t \) to the outputs \( y_t \) as well as weights that link the output
-from the previous time \( y_{t-1} \) and \( y_t \). The figure here shows an
-example of a simple RNN.
-
-More material will be added here.
+
+
# For matrices and calculations
+importnumpyasnp
+# For machine learning (backend for keras)
+importtensorflowastf
+# User-friendly machine learning library
+# Front end for TensorFlow
+importtensorflow.keras
+# Different methods from Keras needed to create an RNN
+# This is not necessary but it shortened function calls
+# that need to be used in the code.
+fromtensorflow.kerasimport datasets, layers, models
+fromtensorflow.keras.layersimport Input
+fromtensorflow.kerasimport regularizers
+fromtensorflow.keras.modelsimport Model, Sequential
+fromtensorflow.keras.layersimport Dense, SimpleRNN, LSTM, GRU
+# For timing the code
+fromtimeitimport default_timer as timer
+# For plotting
+importmatplotlib.pyplotasplt
+
+# The data set
+datatype='VaryDimension'
+X_tot = np.arange(2, 42, 2)
+y_tot = np.array([-0.03077640549, -0.08336233266, -0.1446729567, -0.2116753732, -0.2830637392, -0.3581341341, -0.436462435, -0.5177783846,
+ -0.6019067271, -0.6887363571, -0.7782028952, -0.8702784034, -0.9649652536, -1.062292565, -1.16231451,
+ -1.265109911, -1.370782966, -1.479465113, -1.591317992, -1.70653767])
+
-
Solving differential equations and eigenvalue problems with RNNs
+
Formatting the Data
-In our discussions of ordinary differential equations and partial
-differential equations using neural networks. Here we will discuss how
-we can solve say ordinary differential equations and eigenvalue
-problems using RNNs. Eigenvalue problems can be solved using RNNs by
-rewriting such a problems as a non-linear differential equation.
+The way the recurrent neural networks are trained in this program
+differs from how machine learning algorithms are usually trained.
+Typically a machine learning algorithm is trained by learning the
+relationship between the x data and the y data. In this program, the
+recurrent neural network will be trained to recognize the relationship
+in a sequence of y values. This is type of data formatting is
+typically used time series forcasting, but it can also be used in any
+extrapolation (time series forecasting is just a specific type of
+extrapolation along the time axis). This method of data formatting
+does not use the x data and assumes that the y data are evenly spaced.
-Instead of starting with a well-known ordinary differential equation,
-we start directly with an eigenvaule problem.
+For a standard machine learning algorithm, the training data has the
+form of (x,y) so the machine learning algorithm learns to assiciate a
+y value with a given x value. This is useful when the test data has x
+values within the same range as the training data. However, for this
+application, the x values of the test data are outside of the x values
+of the training data and the traditional method of training a machine
+learning algorithm does not work as well. For this reason, the
+recurrent neural network is trained on sequences of y values of the
+form ((y1, y2), y3), so that the network is concerned with learning
+the pattern of the y data and not the relation between the x and y
+data. As long as the pattern of y data outside of the training region
+stays relatively stable compared to what was inside the training
+region, this method of training can produce accurate extrapolations to
+y values far removed from the training data set.
+
+
+
+
+
+
+
+
+
+
+
+
# FORMAT_DATA
+defformat_data(data, length_of_sequence =2):
+ """
+ Inputs:
+ data(a numpy array): the data that will be the inputs to the recurrent neural
+ network
+ length_of_sequence (an int): the number of elements in one iteration of the
+ sequence patter. For a function approximator use length_of_sequence = 2.
+ Returns:
+ rnn_input (a 3D numpy array): the input data for the recurrent neural network. Its
+ dimensions are length of data - length of sequence, length of sequence,
+ dimnsion of data
+ rnn_output (a numpy array): the training data for the neural network
+ Formats data to be used in a recurrent neural network.
+ """
+
+ X, Y = [], []
+ for i inrange(len(data)-length_of_sequence):
+ # Get the next length_of_sequence elements
+ a = data[i:i+length_of_sequence]
+ # Get the element that immediately follows that
+ b = data[i+length_of_sequence]
+ # Reshape so that each data point is contained in its own array
+ a = np.reshape (a, (len(a), 1))
+ X.append(a)
+ Y.append(b)
+ rnn_input = np.array(X)
+ rnn_output = np.array(Y)
+
+ return rnn_input, rnn_output
+
+
+# ## Defining the Recurrent Neural Network Using Keras
+#
+# The following method defines a simple recurrent neural network in keras consisting of one input layer, one hidden layer, and one output layer.
+
+defrnn(length_of_sequences, batch_size =None, stateful =False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with one hidden layer and returns the model.
+ """
+ # Number of neurons in the input and output layers
+ in_out_neurons =1
+ # Number of neurons in the hidden layer
+ hidden_neurons =200
+ # Define the input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Define the hidden layer as a simple RNN layer with a set number of neurons and add it to
+ # the network immediately after the input layer
+ rnn = SimpleRNN(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN")(inp)
+ # Define the output layer as a dense neural network layer (standard neural network layer)
+ #and add it to the network immediately after the hidden layer.
+ dens = Dense(in_out_neurons,name="dense")(rnn)
+ # Create the machine learning model starting with the input layer and ending with the
+ # output layer
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the machine learning model using the mean squared error function as the loss
+ # function and an Adams optimizer.
+ model.compile(loss="mean_squared_error", optimizer="adam")
+ return model
+
-
Long-Short Time Memory
+
Predicting New Points With A Trained Recurrent Neural Network
-Discussions about dynamic unrolling through time. discuss memory cells, input and output
+
+
deftest_rnn (x1, y_test, plot_min, plot_max):
+ """
+ Inputs:
+ x1 (a list or numpy array): The complete x component of the data set
+ y_test (a list or numpy array): The complete y component of the data set
+ plot_min (an int or float): the smallest x value used in the training data
+ plot_max (an int or float): the largest x valye used in the training data
+ Returns:
+ None.
+ Uses a trained recurrent neural network model to predict future points in the
+ series. Computes the MSE of the predicted data set from the true data set, saves
+ the predicted data set to a csv file, and plots the predicted and true data sets w
+ while also displaying the data range used for training.
+ """
+ # Add the training data as the first dim points in the predicted data array as these
+ # are known values.
+ y_pred = y_test[:dim].tolist()
+ # Generate the first input to the trained recurrent neural network using the last two
+ # points of the training data. Based on how the network was trained this means that it
+ # will predict the first point in the data set after the training data. All of the
+ # brackets are necessary for Tensorflow.
+ next_input = np.array([[[y_test[dim-2]], [y_test[dim-1]]]])
+ # Save the very last point in the training data set. This will be used later.
+ last = [y_test[dim-1]]
+
+ # Iterate until the complete data set is created.
+ for i inrange (dim, len(y_test)):
+ # Predict the next point in the data set using the previous two points.
+ next= model.predict(next_input)
+ # Append just the number of the predicted data set
+ y_pred.append(next[0][0])
+ # Create the input that will be used to predict the next data point in the data set.
+ next_input = np.array([[last, next[0]]], dtype=np.float64)
+ last =next
+
+ # Print the mean squared error between the known data set and the predicted data set.
+ print('MSE: ', np.square(np.subtract(y_test, y_pred)).mean())
+ # Save the predicted data set as a csv file for later use
+ name = datatype +'Predicted'+str(dim)+'.csv'
+ np.savetxt(name, y_pred, delimiter=',')
+ # Plot the known data set and the predicted data set. The red box represents the region that was used
+ # for the training data.
+ fig, ax = plt.subplots()
+ ax.plot(x1, y_test, label="true", linewidth=3)
+ ax.plot(x1, y_pred, 'g-.',label="predicted", linewidth=4)
+ ax.legend()
+ # Created a red region to represent the points used in the training data.
+ ax.axvspan(plot_min, plot_max, alpha=0.25, color='red')
+ plt.show()
+
+# Check to make sure the data set is complete
+assertlen(X_tot) ==len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+model = rnn(length_of_sequences = rnn_input.shape[1])
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
-
Autoencoders: Overarching view
+
Other Things to Try
-Autoencoders are artificial neural networks capable of learning
-efficient representations of the input data (these representations are called codings) without
-any supervision (i.e., the training set is unlabeled). These codings
-typically have a much lower dimensionality than the input data, making
-autoencoders useful for dimensionality reduction.
+Changing the size of the recurrent neural network and its parameters
+can drastically change the results you get from the model. The below
+code takes the simple recurrent neural network from above and adds a
+second hidden layer, changes the number of neurons in the hidden
+layer, and explicitly declares the activation function of the hidden
+layers to be a sigmoid function. The loss function and optimizer can
+also be changed but are kept the same as the above network. These
+parameters can be tuned to provide the optimal result from the
+network. For some ideas on how to improve the performance of a
+recurrent neural network.
-More importantly, autoencoders act as powerful feature detectors, and
-they can be used for unsupervised pretraining of deep neural networks.
-
-Lastly, they are capable of randomly generating new data that looks
-very similar to the training data; this is called a generative
-model. For example, you could train an autoencoder on pictures of
-faces, and it would then be able to generate new faces. Surprisingly,
-autoencoders work by simply learning to copy their inputs to their
-outputs. This may sound like a trivial task, but we will see that
-constraining the network in various ways can make it rather
-difficult. For example, you can limit the size of the internal
-representation, or you can add noise to the inputs and train the
-network to recover the original inputs. These constraints prevent the
-autoencoder from trivially copying the inputs directly to the outputs,
-which forces it to learn efficient ways of representing the data. In
-short, the codings are byproducts of the autoencoder’s attempt to
-learn the identity function under some constraints.
+
+
defrnn_2layers(length_of_sequences, batch_size =None, stateful =False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with two hidden layers and returns the model.
+ """
+ # Number of neurons in the input and output layers
+ in_out_neurons =1
+ # Number of neurons in the hidden layer, increased from the first network
+ hidden_neurons =500
+ # Define the input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Create two hidden layers instead of one hidden layer. Explicitly set the activation
+ # function to be the sigmoid function (the default value is hyperbolic tangent)
+ rnn1 = SimpleRNN(hidden_neurons,
+ return_sequences=True, # This needs to be True if another hidden layer is to follow
+ stateful = stateful, activation ='sigmoid',
+ name="RNN1")(inp)
+ rnn2 = SimpleRNN(hidden_neurons,
+ return_sequences=False, activation ='sigmoid',
+ stateful = stateful,
+ name="RNN2")(rnn1)
+ # Define the output layer as a dense neural network layer (standard neural network layer)
+ #and add it to the network immediately after the hidden layer.
+ dens = Dense(in_out_neurons,name="dense")(rnn2)
+ # Create the machine learning model starting with the input layer and ending with the
+ # output layer
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the machine learning model using the mean squared error function as the loss
+ # function and an Adams optimizer.
+ model.compile(loss="mean_squared_error", optimizer="adam")
+ return model
+# Check to make sure the data set is complete
+assertlen(X_tot) ==len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+model = rnn_2layers(length_of_sequences =2)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
+The first network created below is similar to the previous network,
+but it replaces the SimpleRNN layers with LSTM layers. The second
+network below has two hidden layers made up of GRUs, which are
+preceeded by two dense (feeddorward) neural network layers. These
+dense layers "preprocess" the data before it reaches the recurrent
+layers. This architecture has been shown to improve the performance
+of recurrent neural networks (see the link above and also
+https://arxiv.org/pdf/1807.02857.pdf.
+
+
+
+
+
deflstm_2layers(length_of_sequences, batch_size =None, stateful =False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with two LSTM hidden layers and returns the model.
+ """
+ # Number of neurons on the input/output layer and the number of neurons in the hidden layer
+ in_out_neurons =1
+ hidden_neurons =250
+ # Input Layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Hidden layers (in this case they are LSTM layers instead if SimpleRNN layers)
+ rnn= LSTM(hidden_neurons,
+ return_sequences=True,
+ stateful = stateful,
+ name="RNN", use_bias=True, activation='tanh')(inp)
+ rnn1 = LSTM(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN1", use_bias=True, activation='tanh')(rnn)
+ # Output layer
+ dens = Dense(in_out_neurons,name="dense")(rnn1)
+ # Define the midel
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the model
+ model.compile(loss='mean_squared_error', optimizer='adam')
+ # Return the model
+ return model
+
+defdnn2_gru2(length_of_sequences, batch_size =None, stateful =False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with four hidden layers (two dense followed by
+ two GRU layers) and returns the model.
+ """
+ # Number of neurons on the input/output layers and hidden layers
+ in_out_neurons =1
+ hidden_neurons =250
+ # Input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Hidden Dense (feedforward) layers
+ dnn = Dense(hidden_neurons/2, activation='relu', name='dnn')(inp)
+ dnn1 = Dense(hidden_neurons/2, activation='relu', name='dnn1')(dnn)
+ # Hidden GRU layers
+ rnn1 = GRU(hidden_neurons,
+ return_sequences=True,
+ stateful = stateful,
+ name="RNN1", use_bias=True)(dnn1)
+ rnn = GRU(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN", use_bias=True)(rnn1)
+ # Output layer
+ dens = Dense(in_out_neurons,name="dense")(rnn)
+ # Define the model
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the mdoel
+ model.compile(loss='mean_squared_error', optimizer='adam')
+ # Return the model
+ return model
+
+# Check to make sure the data set is complete
+assertlen(X_tot) ==len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+# Change the method name to reflect which network you want to use
+model = dnn2_gru2(length_of_sequences =2)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
+
+# ### Training Recurrent Neural Networks in the Standard Way (i.e. learning the relationship between the X and Y data)
+#
+# Finally, comparing the performace of a recurrent neural network using the standard data formatting to the performance of the network with time sequence data formatting shows the benefit of this type of data formatting with extrapolation.
+
+# Check to make sure the data set is complete
+assertlen(X_tot) ==len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+# Reshape the data for Keras specifications
+X_train = X_train.reshape((dim, 1))
+y_train = y_train.reshape((dim, 1))
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+# Set the sequence length to 1 for regular data formatting
+model = rnn(length_of_sequences =1)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(X_train, y_train, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict the remaining data points
+X_pred = X_tot[dim:]
+X_pred = X_pred.reshape((len(X_pred), 1))
+y_model = model.predict(X_pred)
+y_pred = np.concatenate((y_tot[:dim], y_model.flatten()))
+
+# Plot the known data set and the predicted data set. The red box represents the region that was used
+# for the training data.
+fig, ax = plt.subplots()
+ax.plot(X_tot, y_tot, label="true", linewidth=3)
+ax.plot(X_tot, y_pred, 'g-.',label="predicted", linewidth=4)
+ax.legend()
+# Created a red region to represent the points used in the training data.
+ax.axvspan(X_tot[0], X_tot[dim], alpha=0.25, color='red')
+plt.show()
+
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
diff --git a/doc/pub/week42/ipynb/ipynb-week42-src.tar.gz b/doc/pub/week42/ipynb/ipynb-week42-src.tar.gz
index ce4d320a6..7f61649e5 100644
Binary files a/doc/pub/week42/ipynb/ipynb-week42-src.tar.gz and b/doc/pub/week42/ipynb/ipynb-week42-src.tar.gz differ
diff --git a/doc/pub/week42/ipynb/week42.ipynb b/doc/pub/week42/ipynb/week42.ipynb
index 8421a220a..6137ecefa 100644
--- a/doc/pub/week42/ipynb/week42.ipynb
+++ b/doc/pub/week42/ipynb/week42.ipynb
@@ -4,13 +4,13 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "\n",
- "# Week 42 Convolutional and Recurrent Neural Networks and Autoencoders\n",
+ "\n",
+ "# Week 42 Convolutional (CNN) and Recurrent (RNN) Neural Networks\n",
"\n",
" \n",
"**Morten Hjorth-Jensen**, Department of Physics, University of Oslo and Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University\n",
"\n",
- "Date: **Oct 15, 2020**\n",
+ "Date: **Oct 16, 2020**\n",
"\n",
"Copyright 1999-2020, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license\n",
"\n",
@@ -19,12 +19,20 @@
"\n",
"## Plan for week 42\n",
"\n",
- "* Thursday: Convolutional Neural Networks and examples\n",
+ "* Thursday: Convolutional Neural Networks and examples. [Video of Lecture](https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureOctober15.mp4?vrtx=view-as-webpage)\n",
"\n",
- "* Friday: Recurrent Neural Networks and Autoencoders\n",
+ "* Friday: Recurrent Neural Networks. [Video of Lecture](https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureOctober16.mp4?vrtx=view-as-webpage)\n",
"\n",
"Reading suggestions for both days: [Aurelien Geron's chapters 13 and 14](https://github.com/CompPhysics/MachineLearning/blob/master/doc/Textbooks/TensorflowML.pdf). Autoencoders are discussed in chapter 15 of Geron's text.\n",
"\n",
+ "**Excellent lectures on CNNs and RNNs.**\n",
+ "\n",
+ "* [Video on Convolutional Neural Networks from MIT](https://www.youtube.com/watch?v=iaSUYvmCekI&ab_channel=AlexanderAmini)\n",
+ "\n",
+ "* [Video on Recurrent Neural Networks from MIT](https://www.youtube.com/watch?v=SEnXr6v2ifU&ab_channel=AlexanderAmini)\n",
+ "\n",
+ "\n",
+ "\n",
"\n",
"\n",
"\n",
@@ -319,29 +327,10 @@
{
"cell_type": "code",
"execution_count": 1,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "inputs = (n_inputs, pixel_width, pixel_height, depth) = (1797, 8, 8, 1)\n",
- "labels = (n_inputs) = (1797,)\n"
- ]
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"%matplotlib inline\n",
"\n",
@@ -397,7 +386,9 @@
{
"cell_type": "code",
"execution_count": 2,
- "metadata": {},
+ "metadata": {
+ "collapsed": false
+ },
"outputs": [],
"source": [
"from tensorflow.keras import datasets, layers, models\n",
@@ -407,7 +398,7 @@
"from tensorflow.keras import optimizers #This allows using whichever optimiser we want (sgd,adam,RMSprop)\n",
"from tensorflow.keras import regularizers #This allows using whichever regularizer we want (l1,l2,l1_l2)\n",
"from tensorflow.keras.utils import to_categorical #This allows using categorical cross entropy as the cost function\n",
- "#rt Cofrom tensorflow.keras imponv2D\n",
+ "#from tensorflow.keras import Conv2D\n",
"#from tensorflow.keras import MaxPooling2D\n",
"#from tensorflow.keras import Flatten\n",
"\n",
@@ -435,7 +426,9 @@
{
"cell_type": "code",
"execution_count": 3,
- "metadata": {},
+ "metadata": {
+ "collapsed": false
+ },
"outputs": [],
"source": [
"def create_convolutional_neural_network_keras(input_shape, receptive_field,\n",
@@ -476,260 +469,10 @@
{
"cell_type": "code",
"execution_count": 4,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "360/360 [==============================] - 0s 361us/sample - loss: 3.4180 - accuracy: 0.1778\n",
- "Learning rate = 1e-05\n",
- "Lambda = 1e-05\n",
- "Test accuracy: 0.178\n",
- "\n",
- "360/360 [==============================] - 0s 320us/sample - loss: 3.4203 - accuracy: 0.0917\n",
- "Learning rate = 1e-05\n",
- "Lambda = 0.0001\n",
- "Test accuracy: 0.092\n",
- "\n",
- "360/360 [==============================] - 0s 316us/sample - loss: 2.7661 - accuracy: 0.1556\n",
- "Learning rate = 1e-05\n",
- "Lambda = 0.001\n",
- "Test accuracy: 0.156\n",
- "\n",
- "360/360 [==============================] - 0s 314us/sample - loss: 3.5947 - accuracy: 0.1167\n",
- "Learning rate = 1e-05\n",
- "Lambda = 0.01\n",
- "Test accuracy: 0.117\n",
- "\n",
- "360/360 [==============================] - 0s 323us/sample - loss: 12.5511 - accuracy: 0.1111\n",
- "Learning rate = 1e-05\n",
- "Lambda = 0.1\n",
- "Test accuracy: 0.111\n",
- "\n",
- "360/360 [==============================] - 0s 293us/sample - loss: 91.5551 - accuracy: 0.2222\n",
- "Learning rate = 1e-05\n",
- "Lambda = 1.0\n",
- "Test accuracy: 0.222\n",
- "\n",
- "360/360 [==============================] - 0s 312us/sample - loss: 518.1064 - accuracy: 0.1889\n",
- "Learning rate = 1e-05\n",
- "Lambda = 10.0\n",
- "Test accuracy: 0.189\n",
- "\n",
- "360/360 [==============================] - 0s 312us/sample - loss: 1.4667 - accuracy: 0.5444\n",
- "Learning rate = 0.0001\n",
- "Lambda = 1e-05\n",
- "Test accuracy: 0.544\n",
- "\n",
- "360/360 [==============================] - 0s 321us/sample - loss: 1.0560 - accuracy: 0.6806\n",
- "Learning rate = 0.0001\n",
- "Lambda = 0.0001\n",
- "Test accuracy: 0.681\n",
- "\n",
- "360/360 [==============================] - 0s 423us/sample - loss: 2.0023 - accuracy: 0.3611\n",
- "Learning rate = 0.0001\n",
- "Lambda = 0.001\n",
- "Test accuracy: 0.361\n",
- "\n",
- "360/360 [==============================] - 0s 325us/sample - loss: 2.5625 - accuracy: 0.4722\n",
- "Learning rate = 0.0001\n",
- "Lambda = 0.01\n",
- "Test accuracy: 0.472\n",
- "\n",
- "360/360 [==============================] - 0s 311us/sample - loss: 10.3306 - accuracy: 0.5694\n",
- "Learning rate = 0.0001\n",
- "Lambda = 0.1\n",
- "Test accuracy: 0.569\n",
- "\n",
- "360/360 [==============================] - 0s 364us/sample - loss: 53.8126 - accuracy: 0.6194\n",
- "Learning rate = 0.0001\n",
- "Lambda = 1.0\n",
- "Test accuracy: 0.619\n",
- "\n",
- "360/360 [==============================] - 0s 304us/sample - loss: 4.5992 - accuracy: 0.0889\n",
- "Learning rate = 0.0001\n",
- "Lambda = 10.0\n",
- "Test accuracy: 0.089\n",
- "\n",
- "360/360 [==============================] - 0s 313us/sample - loss: 0.2762 - accuracy: 0.9194\n",
- "Learning rate = 0.001\n",
- "Lambda = 1e-05\n",
- "Test accuracy: 0.919\n",
- "\n",
- "360/360 [==============================] - 0s 318us/sample - loss: 0.2421 - accuracy: 0.9417\n",
- "Learning rate = 0.001\n",
- "Lambda = 0.0001\n",
- "Test accuracy: 0.942\n",
- "\n",
- "360/360 [==============================] - 0s 305us/sample - loss: 0.3346 - accuracy: 0.9278\n",
- "Learning rate = 0.001\n",
- "Lambda = 0.001\n",
- "Test accuracy: 0.928\n",
- "\n",
- "360/360 [==============================] - 0s 295us/sample - loss: 1.1583 - accuracy: 0.9194\n",
- "Learning rate = 0.001\n",
- "Lambda = 0.01\n",
- "Test accuracy: 0.919\n",
- "\n",
- "360/360 [==============================] - 0s 308us/sample - loss: 5.7769 - accuracy: 0.9194\n",
- "Learning rate = 0.001\n",
- "Lambda = 0.1\n",
- "Test accuracy: 0.919\n",
- "\n",
- "360/360 [==============================] - 0s 317us/sample - loss: 2.5828 - accuracy: 0.1472\n",
- "Learning rate = 0.001\n",
- "Lambda = 1.0\n",
- "Test accuracy: 0.147\n",
- "\n",
- "360/360 [==============================] - 0s 341us/sample - loss: 2.3034 - accuracy: 0.0778\n",
- "Learning rate = 0.001\n",
- "Lambda = 10.0\n",
- "Test accuracy: 0.078\n",
- "\n",
- "360/360 [==============================] - 0s 313us/sample - loss: 0.0683 - accuracy: 0.9694\n",
- "Learning rate = 0.01\n",
- "Lambda = 1e-05\n",
- "Test accuracy: 0.969\n",
- "\n",
- "360/360 [==============================] - 0s 352us/sample - loss: 0.0946 - accuracy: 0.9694\n",
- "Learning rate = 0.01\n",
- "Lambda = 0.0001\n",
- "Test accuracy: 0.969\n",
- "\n",
- "360/360 [==============================] - 0s 325us/sample - loss: 0.1989 - accuracy: 0.9694\n",
- "Learning rate = 0.01\n",
- "Lambda = 0.001\n",
- "Test accuracy: 0.969\n",
- "\n",
- "360/360 [==============================] - 0s 309us/sample - loss: 0.6825 - accuracy: 0.9806\n",
- "Learning rate = 0.01\n",
- "Lambda = 0.01\n",
- "Test accuracy: 0.981\n",
- "\n",
- "360/360 [==============================] - 0s 318us/sample - loss: 0.9712 - accuracy: 0.9306\n",
- "Learning rate = 0.01\n",
- "Lambda = 0.1\n",
- "Test accuracy: 0.931\n",
- "\n",
- "360/360 [==============================] - 0s 323us/sample - loss: 2.3063 - accuracy: 0.0889\n",
- "Learning rate = 0.01\n",
- "Lambda = 1.0\n",
- "Test accuracy: 0.089\n",
- "\n",
- "360/360 [==============================] - 0s 301us/sample - loss: 2.3065 - accuracy: 0.0889\n",
- "Learning rate = 0.01\n",
- "Lambda = 10.0\n",
- "Test accuracy: 0.089\n",
- "\n",
- "360/360 [==============================] - 0s 306us/sample - loss: 0.4443 - accuracy: 0.9056\n",
- "Learning rate = 0.1\n",
- "Lambda = 1e-05\n",
- "Test accuracy: 0.906\n",
- "\n",
- "360/360 [==============================] - 0s 315us/sample - loss: 0.1760 - accuracy: 0.9611\n",
- "Learning rate = 0.1\n",
- "Lambda = 0.0001\n",
- "Test accuracy: 0.961\n",
- "\n",
- "360/360 [==============================] - 0s 307us/sample - loss: 0.2395 - accuracy: 0.9694\n",
- "Learning rate = 0.1\n",
- "Lambda = 0.001\n",
- "Test accuracy: 0.969\n",
- "\n",
- "360/360 [==============================] - 0s 367us/sample - loss: 0.4562 - accuracy: 0.9083\n",
- "Learning rate = 0.1\n",
- "Lambda = 0.01\n",
- "Test accuracy: 0.908\n",
- "\n",
- "360/360 [==============================] - 0s 391us/sample - loss: 1.4405 - accuracy: 0.7972\n",
- "Learning rate = 0.1\n",
- "Lambda = 0.1\n",
- "Test accuracy: 0.797\n",
- "\n",
- "360/360 [==============================] - 0s 318us/sample - loss: 2.3075 - accuracy: 0.0889\n",
- "Learning rate = 0.1\n",
- "Lambda = 1.0\n",
- "Test accuracy: 0.089\n",
- "\n",
- "360/360 [==============================] - 0s 320us/sample - loss: nan - accuracy: 0.0778\n",
- "Learning rate = 0.1\n",
- "Lambda = 10.0\n",
- "Test accuracy: 0.078\n",
- "\n",
- "360/360 [==============================] - 0s 350us/sample - loss: 11.7820 - accuracy: 0.0917\n",
- "Learning rate = 1.0\n",
- "Lambda = 1e-05\n",
- "Test accuracy: 0.092\n",
- "\n",
- "360/360 [==============================] - 0s 315us/sample - loss: 525.7493 - accuracy: 0.0889\n",
- "Learning rate = 1.0\n",
- "Lambda = 0.0001\n",
- "Test accuracy: 0.089\n",
- "\n",
- "360/360 [==============================] - 0s 326us/sample - loss: 14.1148 - accuracy: 0.0778\n",
- "Learning rate = 1.0\n",
- "Lambda = 0.001\n",
- "Test accuracy: 0.078\n",
- "\n",
- "360/360 [==============================] - 0s 330us/sample - loss: 2.3088 - accuracy: 0.1056\n",
- "Learning rate = 1.0\n",
- "Lambda = 0.01\n",
- "Test accuracy: 0.106\n",
- "\n",
- "360/360 [==============================] - 0s 315us/sample - loss: 2.3101 - accuracy: 0.1056\n",
- "Learning rate = 1.0\n",
- "Lambda = 0.1\n",
- "Test accuracy: 0.106\n",
- "\n",
- "360/360 [==============================] - 0s 309us/sample - loss: nan - accuracy: 0.0778\n",
- "Learning rate = 1.0\n",
- "Lambda = 1.0\n",
- "Test accuracy: 0.078\n",
- "\n",
- "360/360 [==============================] - 0s 322us/sample - loss: nan - accuracy: 0.0778\n",
- "Learning rate = 1.0\n",
- "Lambda = 10.0\n",
- "Test accuracy: 0.078\n",
- "\n",
- "360/360 [==============================] - 0s 297us/sample - loss: 6288412.0000 - accuracy: 0.1056\n",
- "Learning rate = 10.0\n",
- "Lambda = 1e-05\n",
- "Test accuracy: 0.106\n",
- "\n",
- "360/360 [==============================] - 0s 483us/sample - loss: 285465.2687 - accuracy: 0.1056\n",
- "Learning rate = 10.0\n",
- "Lambda = 0.0001\n",
- "Test accuracy: 0.106\n",
- "\n",
- "360/360 [==============================] - 0s 348us/sample - loss: 2.3793 - accuracy: 0.1250\n",
- "Learning rate = 10.0\n",
- "Lambda = 0.001\n",
- "Test accuracy: 0.125\n",
- "\n",
- "360/360 [==============================] - 0s 318us/sample - loss: 2.3507 - accuracy: 0.1056\n",
- "Learning rate = 10.0\n",
- "Lambda = 0.01\n",
- "Test accuracy: 0.106\n",
- "\n",
- "360/360 [==============================] - 0s 638us/sample - loss: nan - accuracy: 0.0778\n",
- "Learning rate = 10.0\n",
- "Lambda = 0.1\n",
- "Test accuracy: 0.078\n",
- "\n",
- "360/360 [==============================] - 0s 432us/sample - loss: nan - accuracy: 0.0778\n",
- "Learning rate = 10.0\n",
- "Lambda = 1.0\n",
- "Test accuracy: 0.078\n",
- "\n",
- "360/360 [==============================] - 0s 318us/sample - loss: nan - accuracy: 0.0778\n",
- "Learning rate = 10.0\n",
- "Lambda = 10.0\n",
- "Test accuracy: 0.078\n",
- "\n"
- ]
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"CNN_keras = np.zeros((len(eta_vals), len(lmbd_vals)), dtype=object)\n",
" \n",
@@ -758,140 +501,11 @@
},
{
"cell_type": "code",
- "execution_count": 6,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "1437/1437 [==============================] - 0s 43us/sample - loss: 3.3022 - accuracy: 0.1872\n",
- "360/360 [==============================] - 0s 121us/sample - loss: 3.4180 - accuracy: 0.1778\n",
- "1437/1437 [==============================] - 0s 86us/sample - loss: 3.3955 - accuracy: 0.1093\n",
- "360/360 [==============================] - 0s 142us/sample - loss: 3.4203 - accuracy: 0.0917\n",
- "1437/1437 [==============================] - 0s 80us/sample - loss: 2.7250 - accuracy: 0.1587\n",
- "360/360 [==============================] - 0s 216us/sample - loss: 2.7661 - accuracy: 0.1556\n",
- "1437/1437 [==============================] - 0s 66us/sample - loss: 3.5698 - accuracy: 0.1343\n",
- "360/360 [==============================] - 0s 46us/sample - loss: 3.5947 - accuracy: 0.1167\n",
- "1437/1437 [==============================] - 0s 63us/sample - loss: 12.5837 - accuracy: 0.0946\n",
- "360/360 [==============================] - 0s 60us/sample - loss: 12.5511 - accuracy: 0.1111\n",
- "1437/1437 [==============================] - 0s 59us/sample - loss: 91.5210 - accuracy: 0.2408\n",
- "360/360 [==============================] - 0s 53us/sample - loss: 91.5551 - accuracy: 0.2222\n",
- "1437/1437 [==============================] - 0s 64us/sample - loss: 518.1178 - accuracy: 0.1969\n",
- "360/360 [==============================] - 0s 48us/sample - loss: 518.1064 - accuracy: 0.1889\n",
- "1437/1437 [==============================] - 0s 66us/sample - loss: 1.4465 - accuracy: 0.5623\n",
- "360/360 [==============================] - 0s 37us/sample - loss: 1.4667 - accuracy: 0.5444\n",
- "1437/1437 [==============================] - 0s 63us/sample - loss: 1.0335 - accuracy: 0.7015\n",
- "360/360 [==============================] - 0s 75us/sample - loss: 1.0560 - accuracy: 0.6806\n",
- "1437/1437 [==============================] - 0s 62us/sample - loss: 1.9454 - accuracy: 0.3730\n",
- "360/360 [==============================] - 0s 48us/sample - loss: 2.0023 - accuracy: 0.3611\n",
- "1437/1437 [==============================] - 0s 92us/sample - loss: 2.4747 - accuracy: 0.5080\n",
- "360/360 [==============================] - 0s 49us/sample - loss: 2.5625 - accuracy: 0.4722\n",
- "1437/1437 [==============================] - 0s 65us/sample - loss: 10.2878 - accuracy: 0.5887\n",
- "360/360 [==============================] - 0s 87us/sample - loss: 10.3306 - accuracy: 0.5694\n",
- "1437/1437 [==============================] - 0s 60us/sample - loss: 53.7810 - accuracy: 0.6548\n",
- "360/360 [==============================] - 0s 45us/sample - loss: 53.8126 - accuracy: 0.6194\n",
- "1437/1437 [==============================] - 0s 55us/sample - loss: 4.5991 - accuracy: 0.1058\n",
- "360/360 [==============================] - 0s 45us/sample - loss: 4.5992 - accuracy: 0.0889\n",
- "1437/1437 [==============================] - 0s 66us/sample - loss: 0.2035 - accuracy: 0.9457\n",
- "360/360 [==============================] - 0s 43us/sample - loss: 0.2762 - accuracy: 0.9194\n",
- "1437/1437 [==============================] - 0s 82us/sample - loss: 0.1869 - accuracy: 0.9617\n",
- "360/360 [==============================] - 0s 46us/sample - loss: 0.2421 - accuracy: 0.9417\n",
- "1437/1437 [==============================] - 0s 59us/sample - loss: 0.2736 - accuracy: 0.9527\n",
- "360/360 [==============================] - 0s 57us/sample - loss: 0.3346 - accuracy: 0.9278\n",
- "1437/1437 [==============================] - 0s 63us/sample - loss: 1.0958 - accuracy: 0.9499\n",
- "360/360 [==============================] - 0s 48us/sample - loss: 1.1583 - accuracy: 0.9194\n",
- "1437/1437 [==============================] - ETA: 0s - loss: 5.7025 - accuracy: 0.95 - 0s 68us/sample - loss: 5.7254 - accuracy: 0.9506\n",
- "360/360 [==============================] - 0s 226us/sample - loss: 5.7769 - accuracy: 0.9194\n",
- "1437/1437 [==============================] - 0s 61us/sample - loss: 2.5815 - accuracy: 0.1886\n",
- "360/360 [==============================] - 0s 56us/sample - loss: 2.5828 - accuracy: 0.1472\n",
- "1437/1437 [==============================] - 0s 61us/sample - loss: 2.3024 - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 39us/sample - loss: 2.3034 - accuracy: 0.0778\n",
- "1437/1437 [==============================] - 0s 68us/sample - loss: 0.0132 - accuracy: 1.0000\n",
- "360/360 [==============================] - 0s 59us/sample - loss: 0.0683 - accuracy: 0.9694\n",
- "1437/1437 [==============================] - 0s 71us/sample - loss: 0.0264 - accuracy: 0.9993\n",
- "360/360 [==============================] - 0s 46us/sample - loss: 0.0946 - accuracy: 0.9694\n",
- "1437/1437 [==============================] - 0s 56us/sample - loss: 0.1294 - accuracy: 0.9944\n",
- "360/360 [==============================] - 0s 55us/sample - loss: 0.1989 - accuracy: 0.9694\n",
- "1437/1437 [==============================] - 0s 67us/sample - loss: 0.6352 - accuracy: 0.9972\n",
- "360/360 [==============================] - 0s 41us/sample - loss: 0.6825 - accuracy: 0.9806\n",
- "1437/1437 [==============================] - 0s 75us/sample - loss: 0.9284 - accuracy: 0.9617\n",
- "360/360 [==============================] - 0s 40us/sample - loss: 0.9712 - accuracy: 0.9306\n",
- "1437/1437 [==============================] - 0s 54us/sample - loss: 2.3020 - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 73us/sample - loss: 2.3063 - accuracy: 0.0889\n",
- "1437/1437 [==============================] - 0s 72us/sample - loss: 2.3020 - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 43us/sample - loss: 2.3065 - accuracy: 0.0889\n",
- "1437/1437 [==============================] - 0s 78us/sample - loss: 0.0163 - accuracy: 0.9972\n",
- "360/360 [==============================] - 0s 56us/sample - loss: 0.4443 - accuracy: 0.9056\n",
- "1437/1437 [==============================] - 0s 60us/sample - loss: 0.0128 - accuracy: 1.0000\n",
- "360/360 [==============================] - 0s 43us/sample - loss: 0.1760 - accuracy: 0.9611\n",
- "1437/1437 [==============================] - 0s 62us/sample - loss: 0.0952 - accuracy: 1.0000\n",
- "360/360 [==============================] - 0s 41us/sample - loss: 0.2395 - accuracy: 0.9694\n",
- "1437/1437 [==============================] - 0s 72us/sample - loss: 0.4259 - accuracy: 0.9179\n",
- "360/360 [==============================] - 0s 52us/sample - loss: 0.4562 - accuracy: 0.9083\n",
- "1437/1437 [==============================] - 0s 63us/sample - loss: 1.3861 - accuracy: 0.8386\n",
- "360/360 [==============================] - 0s 50us/sample - loss: 1.4405 - accuracy: 0.7972\n",
- "1437/1437 [==============================] - 0s 62us/sample - loss: 2.3020 - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 43us/sample - loss: 2.3075 - accuracy: 0.0889\n",
- "1437/1437 [==============================] - 0s 65us/sample - loss: nan - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 45us/sample - loss: nan - accuracy: 0.0778\n",
- "1437/1437 [==============================] - 0s 68us/sample - loss: 11.7729 - accuracy: 0.1016\n",
- "360/360 [==============================] - 0s 53us/sample - loss: 11.7820 - accuracy: 0.0917\n",
- "1437/1437 [==============================] - 0s 54us/sample - loss: 525.7408 - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 54us/sample - loss: 525.7493 - accuracy: 0.0889\n",
- "1437/1437 [==============================] - 0s 54us/sample - loss: 14.1025 - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 51us/sample - loss: 14.1148 - accuracy: 0.0778\n",
- "1437/1437 [==============================] - 0s 47us/sample - loss: 2.3039 - accuracy: 0.0995\n",
- "360/360 [==============================] - 0s 44us/sample - loss: 2.3088 - accuracy: 0.1056\n",
- "1437/1437 [==============================] - 0s 61us/sample - loss: 2.3031 - accuracy: 0.0995\n",
- "360/360 [==============================] - 0s 41us/sample - loss: 2.3101 - accuracy: 0.1056\n",
- "1437/1437 [==============================] - 0s 79us/sample - loss: nan - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 36us/sample - loss: nan - accuracy: 0.0778\n",
- "1437/1437 [==============================] - 0s 58us/sample - loss: nan - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 61us/sample - loss: nan - accuracy: 0.0778\n",
- "1437/1437 [==============================] - 0s 60us/sample - loss: 6288412.0000 - accuracy: 0.0995\n",
- "360/360 [==============================] - 0s 62us/sample - loss: 6288412.0000 - accuracy: 0.1056\n",
- "1437/1437 [==============================] - 0s 67us/sample - loss: 285465.2778 - accuracy: 0.0995\n"
- ]
- },
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "360/360 [==============================] - 0s 49us/sample - loss: 285465.2687 - accuracy: 0.1056\n",
- "1437/1437 [==============================] - 0s 92us/sample - loss: 2.3807 - accuracy: 0.0953\n",
- "360/360 [==============================] - 0s 53us/sample - loss: 2.3793 - accuracy: 0.1250\n",
- "1437/1437 [==============================] - 0s 55us/sample - loss: 2.3769 - accuracy: 0.0995\n",
- "360/360 [==============================] - 0s 58us/sample - loss: 2.3507 - accuracy: 0.1056\n",
- "1437/1437 [==============================] - 0s 57us/sample - loss: nan - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 41us/sample - loss: nan - accuracy: 0.0778\n",
- "1437/1437 [==============================] - 0s 64us/sample - loss: nan - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 54us/sample - loss: nan - accuracy: 0.0778\n",
- "1437/1437 [==============================] - 0s 64us/sample - loss: nan - accuracy: 0.1044\n",
- "360/360 [==============================] - 0s 231us/sample - loss: nan - accuracy: 0.0778\n"
- ]
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
+ "execution_count": 5,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"# visual representation of grid search\n",
"# uses seaborn heatmap, could probably do this in matplotlib\n",
@@ -939,18 +553,11 @@
},
{
"cell_type": "code",
- "execution_count": 7,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Downloading data from https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz\n",
- "170500096/170498071 [==============================] - 35s 0us/step\n"
- ]
- }
- ],
+ "execution_count": 6,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"import tensorflow as tf\n",
"\n",
@@ -975,23 +582,15 @@
},
{
"cell_type": "code",
- "execution_count": 9,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
+ "execution_count": 7,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"class_names = ['airplane', 'automobile', 'bird', 'cat', 'deer',\n",
" 'dog', 'frog', 'horse', 'ship', 'truck']\n",
+ "​\n",
"plt.figure(figsize=(10,10))\n",
"for i in range(25):\n",
" plt.subplot(5,5,i+1)\n",
@@ -1018,34 +617,11 @@
},
{
"cell_type": "code",
- "execution_count": 10,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Model: \"sequential_49\"\n",
- "_________________________________________________________________\n",
- "Layer (type) Output Shape Param # \n",
- "=================================================================\n",
- "conv2d_49 (Conv2D) (None, 30, 30, 32) 896 \n",
- "_________________________________________________________________\n",
- "max_pooling2d_49 (MaxPooling (None, 15, 15, 32) 0 \n",
- "_________________________________________________________________\n",
- "conv2d_50 (Conv2D) (None, 13, 13, 64) 18496 \n",
- "_________________________________________________________________\n",
- "max_pooling2d_50 (MaxPooling (None, 6, 6, 64) 0 \n",
- "_________________________________________________________________\n",
- "conv2d_51 (Conv2D) (None, 4, 4, 64) 36928 \n",
- "=================================================================\n",
- "Total params: 56,320\n",
- "Trainable params: 56,320\n",
- "Non-trainable params: 0\n",
- "_________________________________________________________________\n"
- ]
- }
- ],
+ "execution_count": 8,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"model = models.Sequential()\n",
"model.add(layers.Conv2D(32, (3, 3), activation='relu', input_shape=(32, 32, 3)))\n",
@@ -1081,45 +657,16 @@
},
{
"cell_type": "code",
- "execution_count": 12,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Model: \"sequential_49\"\n",
- "_________________________________________________________________\n",
- "Layer (type) Output Shape Param # \n",
- "=================================================================\n",
- "conv2d_49 (Conv2D) (None, 30, 30, 32) 896 \n",
- "_________________________________________________________________\n",
- "max_pooling2d_49 (MaxPooling (None, 15, 15, 32) 0 \n",
- "_________________________________________________________________\n",
- "conv2d_50 (Conv2D) (None, 13, 13, 64) 18496 \n",
- "_________________________________________________________________\n",
- "max_pooling2d_50 (MaxPooling (None, 6, 6, 64) 0 \n",
- "_________________________________________________________________\n",
- "conv2d_51 (Conv2D) (None, 4, 4, 64) 36928 \n",
- "_________________________________________________________________\n",
- "flatten_49 (Flatten) (None, 1024) 0 \n",
- "_________________________________________________________________\n",
- "dense_98 (Dense) (None, 64) 65600 \n",
- "_________________________________________________________________\n",
- "dense_99 (Dense) (None, 10) 650 \n",
- "=================================================================\n",
- "Total params: 122,570\n",
- "Trainable params: 122,570\n",
- "Non-trainable params: 0\n",
- "_________________________________________________________________\n"
- ]
- }
- ],
+ "execution_count": 9,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"model.add(layers.Flatten())\n",
"model.add(layers.Dense(64, activation='relu'))\n",
"model.add(layers.Dense(10))\n",
- "#Here's the complete architecture of our model.\n",
+ "Here's the complete architecture of our model.\n",
"\n",
"model.summary()"
]
@@ -1135,42 +682,16 @@
},
{
"cell_type": "code",
- "execution_count": 13,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Train on 50000 samples, validate on 10000 samples\n",
- "Epoch 1/10\n",
- "50000/50000 [==============================] - 40s 793us/sample - loss: 1.5115 - accuracy: 0.4515 - val_loss: 1.2411 - val_accuracy: 0.5545\n",
- "Epoch 2/10\n",
- "50000/50000 [==============================] - 41s 826us/sample - loss: 1.1297 - accuracy: 0.6006 - val_loss: 1.0419 - val_accuracy: 0.6307\n",
- "Epoch 3/10\n",
- "50000/50000 [==============================] - 43s 870us/sample - loss: 0.9842 - accuracy: 0.6534 - val_loss: 1.0402 - val_accuracy: 0.6314\n",
- "Epoch 4/10\n",
- "50000/50000 [==============================] - 43s 869us/sample - loss: 0.8824 - accuracy: 0.6894 - val_loss: 0.9944 - val_accuracy: 0.6599\n",
- "Epoch 5/10\n",
- "50000/50000 [==============================] - 40s 803us/sample - loss: 0.8098 - accuracy: 0.7171 - val_loss: 0.9176 - val_accuracy: 0.6829\n",
- "Epoch 6/10\n",
- "50000/50000 [==============================] - 46s 925us/sample - loss: 0.7469 - accuracy: 0.7370 - val_loss: 0.8683 - val_accuracy: 0.7072\n",
- "Epoch 7/10\n",
- "50000/50000 [==============================] - 43s 857us/sample - loss: 0.6939 - accuracy: 0.7546 - val_loss: 0.8628 - val_accuracy: 0.7055\n",
- "Epoch 8/10\n",
- "50000/50000 [==============================] - 38s 770us/sample - loss: 0.6492 - accuracy: 0.7719 - val_loss: 0.8725 - val_accuracy: 0.7120\n",
- "Epoch 9/10\n",
- "50000/50000 [==============================] - 37s 743us/sample - loss: 0.6064 - accuracy: 0.7881 - val_loss: 0.8604 - val_accuracy: 0.7144\n",
- "Epoch 10/10\n",
- "50000/50000 [==============================] - 36s 715us/sample - loss: 0.5675 - accuracy: 0.8003 - val_loss: 0.8882 - val_accuracy: 0.7137\n"
- ]
- }
- ],
+ "execution_count": 10,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"model.compile(optimizer='adam',\n",
" loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
" metrics=['accuracy'])\n",
- "\n",
+ "​\n",
"history = model.fit(train_images, train_labels, epochs=10, \n",
" validation_data=(test_images, test_labels))"
]
@@ -1184,6 +705,7 @@
},
{
"cell_type": "code",
+<<<<<<< HEAD
"execution_count": 14,
"metadata": {},
"outputs": [
@@ -1206,6 +728,13 @@
"output_type": "display_data"
}
],
+=======
+ "execution_count": 11,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
+>>>>>>> b95146b3e7172ba164de8bbc2afe0d9cc4dba4f6
"source": [
"plt.plot(history.history['accuracy'], label='accuracy')\n",
"plt.plot(history.history['val_accuracy'], label = 'val_accuracy')\n",
@@ -1243,11 +772,21 @@
"systems such as automatic translation and speech-to-text.\n",
"\n",
"\n",
+ "\n",
+ "\n",
+ "## Set up of an RNN\n",
+ "\n",
+ "See the [handwritten notes](https://github.com/CompPhysics/MachineLearning/blob/master/doc/HandWrittenNotes/NotesOctober16.pdf) and the [video from the lecture of October 16](https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureOctober16.mp4?vrtx=view-as-webpage).\n",
+ "\n",
+ "More text will be added later.\n",
+ "\n",
+ "\n",
"## A simple example"
]
},
{
"cell_type": "code",
+<<<<<<< HEAD
"execution_count": 15,
"metadata": {},
"outputs": [
@@ -1494,6 +1033,13 @@
"output_type": "display_data"
}
],
+=======
+ "execution_count": 12,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
+>>>>>>> b95146b3e7172ba164de8bbc2afe0d9cc4dba4f6
"source": [
"# Start importing packages\n",
"import pandas as pd\n",
@@ -1569,93 +1115,622 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "## Set up of an RNN\n",
+ "## An extrapolation example\n",
"\n",
- "The figure here displays a simple example of an RNN, with inputs $x_t$\n",
- "at a given time $t$ and outputs $y_t$. Introducing time as a variable\n",
- "offers an intutitive way of understanding these networks. In addition\n",
- "to the inputs $x_t$, the layer at a time $t$ receives also as input\n",
- "the output from the previous layer $t-1$, that is $y_{t1}$.\n",
+ "The following code provides an example of how recurrent neural\n",
+ "networks can be used to extrapolate to unknown values of physics data\n",
+ "sets. Specifically, the data sets used in this program come from\n",
+ "a quantum mechanical many-body calculation of energies as functions of the number of particles."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 13,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
+ "source": [
"\n",
- "This means also that we need to have weights that link both the inputs\n",
- "$x_t$ to the outputs $y_t$ as well as weights that link the output\n",
- "from the previous time $y_{t-1}$ and $y_t$. The figure here shows an\n",
- "example of a simple RNN.\n",
- "\n",
- "More material will be added here.\n",
+ "# For matrices and calculations\n",
+ "import numpy as np\n",
+ "# For machine learning (backend for keras)\n",
+ "import tensorflow as tf\n",
+ "# User-friendly machine learning library\n",
+ "# Front end for TensorFlow\n",
+ "import tensorflow.keras\n",
+ "# Different methods from Keras needed to create an RNN\n",
+ "# This is not necessary but it shortened function calls \n",
+ "# that need to be used in the code.\n",
+ "from tensorflow.keras import datasets, layers, models\n",
+ "from tensorflow.keras.layers import Input\n",
+ "from tensorflow.keras import regularizers\n",
+ "from tensorflow.keras.models import Model, Sequential\n",
+ "from tensorflow.keras.layers import Dense, SimpleRNN, LSTM, GRU\n",
+ "# For timing the code\n",
+ "from timeit import default_timer as timer\n",
+ "# For plotting\n",
+ "import matplotlib.pyplot as plt\n",
"\n",
"\n",
- "## Solving differential equations and eigenvalue problems with RNNs\n",
+ "# The data set\n",
+ "datatype='VaryDimension'\n",
+ "X_tot = np.arange(2, 42, 2)\n",
+ "y_tot = np.array([-0.03077640549, -0.08336233266, -0.1446729567, -0.2116753732, -0.2830637392, -0.3581341341, -0.436462435, -0.5177783846,\n",
+ "\t-0.6019067271, -0.6887363571, -0.7782028952, -0.8702784034, -0.9649652536, -1.062292565, -1.16231451, \n",
+ "\t-1.265109911, -1.370782966, -1.479465113, -1.591317992, -1.70653767])"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Formatting the Data\n",
+ "\n",
+ "The way the recurrent neural networks are trained in this program\n",
+ "differs from how machine learning algorithms are usually trained.\n",
+ "Typically a machine learning algorithm is trained by learning the\n",
+ "relationship between the x data and the y data. In this program, the\n",
+ "recurrent neural network will be trained to recognize the relationship\n",
+ "in a sequence of y values. This is type of data formatting is\n",
+ "typically used time series forcasting, but it can also be used in any\n",
+ "extrapolation (time series forecasting is just a specific type of\n",
+ "extrapolation along the time axis). This method of data formatting\n",
+ "does not use the x data and assumes that the y data are evenly spaced.\n",
+ "\n",
+ "For a standard machine learning algorithm, the training data has the\n",
+ "form of (x,y) so the machine learning algorithm learns to assiciate a\n",
+ "y value with a given x value. This is useful when the test data has x\n",
+ "values within the same range as the training data. However, for this\n",
+ "application, the x values of the test data are outside of the x values\n",
+ "of the training data and the traditional method of training a machine\n",
+ "learning algorithm does not work as well. For this reason, the\n",
+ "recurrent neural network is trained on sequences of y values of the\n",
+ "form ((y1, y2), y3), so that the network is concerned with learning\n",
+ "the pattern of the y data and not the relation between the x and y\n",
+ "data. As long as the pattern of y data outside of the training region\n",
+ "stays relatively stable compared to what was inside the training\n",
+ "region, this method of training can produce accurate extrapolations to\n",
+ "y values far removed from the training data set.\n",
"\n",
"\n",
+ "\n",
+ "\n",
+ "\n",
+ "\n",
+ "\n",
+ ""
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 14,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
+ "source": [
+ "# FORMAT_DATA\n",
+ "def format_data(data, length_of_sequence = 2): \n",
+ " \"\"\"\n",
+ " Inputs:\n",
+ " data(a numpy array): the data that will be the inputs to the recurrent neural\n",
+ " network\n",
+ " length_of_sequence (an int): the number of elements in one iteration of the\n",
+ " sequence patter. For a function approximator use length_of_sequence = 2.\n",
+ " Returns:\n",
+ " rnn_input (a 3D numpy array): the input data for the recurrent neural network. Its\n",
+ " dimensions are length of data - length of sequence, length of sequence, \n",
+ " dimnsion of data\n",
+ " rnn_output (a numpy array): the training data for the neural network\n",
+ " Formats data to be used in a recurrent neural network.\n",
+ " \"\"\"\n",
"\n",
- "In our discussions of ordinary differential equations and partial\n",
- "differential equations using neural networks. Here we will discuss how\n",
- "we can solve say ordinary differential equations and eigenvalue\n",
- "problems using RNNs. Eigenvalue problems can be solved using RNNs by\n",
- "rewriting such a problems as a non-linear differential equation.\n",
+ " X, Y = [], []\n",
+ " for i in range(len(data)-length_of_sequence):\n",
+ " # Get the next length_of_sequence elements\n",
+ " a = data[i:i+length_of_sequence]\n",
+ " # Get the element that immediately follows that\n",
+ " b = data[i+length_of_sequence]\n",
+ " # Reshape so that each data point is contained in its own array\n",
+ " a = np.reshape (a, (len(a), 1))\n",
+ " X.append(a)\n",
+ " Y.append(b)\n",
+ " rnn_input = np.array(X)\n",
+ " rnn_output = np.array(Y)\n",
"\n",
- "Instead of starting with a well-known ordinary differential equation,\n",
- "we start directly with an eigenvaule problem.\n",
+ " return rnn_input, rnn_output\n",
"\n",
"\n",
+ "# ## Defining the Recurrent Neural Network Using Keras\n",
+ "# \n",
+ "# The following method defines a simple recurrent neural network in keras consisting of one input layer, one hidden layer, and one output layer.\n",
"\n",
- "## Long-Short Time Memory\n",
+ "def rnn(length_of_sequences, batch_size = None, stateful = False):\n",
+ " \"\"\"\n",
+ " Inputs:\n",
+ " length_of_sequences (an int): the number of y values in \"x data\". This is determined\n",
+ " when the data is formatted\n",
+ " batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.\n",
+ " stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.\n",
+ " Returns:\n",
+ " model (a Keras model): The recurrent neural network that is built and compiled by this\n",
+ " method\n",
+ " Builds and compiles a recurrent neural network with one hidden layer and returns the model.\n",
+ " \"\"\"\n",
+ " # Number of neurons in the input and output layers\n",
+ " in_out_neurons = 1\n",
+ " # Number of neurons in the hidden layer\n",
+ " hidden_neurons = 200\n",
+ " # Define the input layer\n",
+ " inp = Input(batch_shape=(batch_size, \n",
+ " length_of_sequences, \n",
+ " in_out_neurons)) \n",
+ " # Define the hidden layer as a simple RNN layer with a set number of neurons and add it to \n",
+ " # the network immediately after the input layer\n",
+ " rnn = SimpleRNN(hidden_neurons, \n",
+ " return_sequences=False,\n",
+ " stateful = stateful,\n",
+ " name=\"RNN\")(inp)\n",
+ " # Define the output layer as a dense neural network layer (standard neural network layer)\n",
+ " #and add it to the network immediately after the hidden layer.\n",
+ " dens = Dense(in_out_neurons,name=\"dense\")(rnn)\n",
+ " # Create the machine learning model starting with the input layer and ending with the \n",
+ " # output layer\n",
+ " model = Model(inputs=[inp],outputs=[dens])\n",
+ " # Compile the machine learning model using the mean squared error function as the loss \n",
+ " # function and an Adams optimizer.\n",
+ " model.compile(loss=\"mean_squared_error\", optimizer=\"adam\") \n",
+ " return model"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Predicting New Points With A Trained Recurrent Neural Network"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 15,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
+ "source": [
+ "def test_rnn (x1, y_test, plot_min, plot_max):\n",
+ " \"\"\"\n",
+ " Inputs:\n",
+ " x1 (a list or numpy array): The complete x component of the data set\n",
+ " y_test (a list or numpy array): The complete y component of the data set\n",
+ " plot_min (an int or float): the smallest x value used in the training data\n",
+ " plot_max (an int or float): the largest x valye used in the training data\n",
+ " Returns:\n",
+ " None.\n",
+ " Uses a trained recurrent neural network model to predict future points in the \n",
+ " series. Computes the MSE of the predicted data set from the true data set, saves\n",
+ " the predicted data set to a csv file, and plots the predicted and true data sets w\n",
+ " while also displaying the data range used for training.\n",
+ " \"\"\"\n",
+ " # Add the training data as the first dim points in the predicted data array as these\n",
+ " # are known values.\n",
+ " y_pred = y_test[:dim].tolist()\n",
+ " # Generate the first input to the trained recurrent neural network using the last two \n",
+ " # points of the training data. Based on how the network was trained this means that it\n",
+ " # will predict the first point in the data set after the training data. All of the \n",
+ " # brackets are necessary for Tensorflow.\n",
+ " next_input = np.array([[[y_test[dim-2]], [y_test[dim-1]]]])\n",
+ " # Save the very last point in the training data set. This will be used later.\n",
+ " last = [y_test[dim-1]]\n",
"\n",
- "Discussions about dynamic unrolling through time. discuss memory cells, input and output\n",
+ " # Iterate until the complete data set is created.\n",
+ " for i in range (dim, len(y_test)):\n",
+ " # Predict the next point in the data set using the previous two points.\n",
+ " next = model.predict(next_input)\n",
+ " # Append just the number of the predicted data set\n",
+ " y_pred.append(next[0][0])\n",
+ " # Create the input that will be used to predict the next data point in the data set.\n",
+ " next_input = np.array([[last, next[0]]], dtype=np.float64)\n",
+ " last = next\n",
+ "\n",
+ " # Print the mean squared error between the known data set and the predicted data set.\n",
+ " print('MSE: ', np.square(np.subtract(y_test, y_pred)).mean())\n",
+ " # Save the predicted data set as a csv file for later use\n",
+ " name = datatype + 'Predicted'+str(dim)+'.csv'\n",
+ " np.savetxt(name, y_pred, delimiter=',')\n",
+ " # Plot the known data set and the predicted data set. The red box represents the region that was used\n",
+ " # for the training data.\n",
+ " fig, ax = plt.subplots()\n",
+ " ax.plot(x1, y_test, label=\"true\", linewidth=3)\n",
+ " ax.plot(x1, y_pred, 'g-.',label=\"predicted\", linewidth=4)\n",
+ " ax.legend()\n",
+ " # Created a red region to represent the points used in the training data.\n",
+ " ax.axvspan(plot_min, plot_max, alpha=0.25, color='red')\n",
+ " plt.show()\n",
+ "\n",
+ "# Check to make sure the data set is complete\n",
+ "assert len(X_tot) == len(y_tot)\n",
+ "\n",
+ "# This is the number of points that will be used in as the training data\n",
+ "dim=12\n",
+ "\n",
+ "# Separate the training data from the whole data set\n",
+ "X_train = X_tot[:dim]\n",
+ "y_train = y_tot[:dim]\n",
"\n",
"\n",
+ "# Generate the training data for the RNN, using a sequence of 2\n",
+ "rnn_input, rnn_training = format_data(y_train, 2)\n",
"\n",
"\n",
- "## Autoencoders: Overarching view\n",
+ "# Create a recurrent neural network in Keras and produce a summary of the \n",
+ "# machine learning model\n",
+ "model = rnn(length_of_sequences = rnn_input.shape[1])\n",
+ "model.summary()\n",
"\n",
- "Autoencoders are artificial neural networks capable of learning\n",
- "efficient representations of the input data (these representations are called codings) without\n",
- "any supervision (i.e., the training set is unlabeled). These codings\n",
- "typically have a much lower dimensionality than the input data, making\n",
- "autoencoders useful for dimensionality reduction. \n",
+ "# Start the timer. Want to time training+testing\n",
+ "start = timer()\n",
+ "# Fit the model using the training data genenerated above using 150 training iterations and a 5%\n",
+ "# validation split. Setting verbose to True prints information about each training iteration.\n",
+ "hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150, \n",
+ " verbose=True,validation_split=0.05)\n",
"\n",
- "More importantly, autoencoders act as powerful feature detectors, and\n",
- "they can be used for unsupervised pretraining of deep neural networks.\n",
+ "for label in [\"loss\",\"val_loss\"]:\n",
+ " plt.plot(hist.history[label],label=label)\n",
"\n",
- "Lastly, they are capable of randomly generating new data that looks\n",
- "very similar to the training data; this is called a generative\n",
- "model. For example, you could train an autoencoder on pictures of\n",
- "faces, and it would then be able to generate new faces. Surprisingly,\n",
- "autoencoders work by simply learning to copy their inputs to their\n",
- "outputs. This may sound like a trivial task, but we will see that\n",
- "constraining the network in various ways can make it rather\n",
- "difficult. For example, you can limit the size of the internal\n",
- "representation, or you can add noise to the inputs and train the\n",
- "network to recover the original inputs. These constraints prevent the\n",
- "autoencoder from trivially copying the inputs directly to the outputs,\n",
- "which forces it to learn efficient ways of representing the data. In\n",
- "short, the codings are byproducts of the autoencoder’s attempt to\n",
- "learn the identity function under some constraints.\n",
+ "plt.ylabel(\"loss\")\n",
+ "plt.xlabel(\"epoch\")\n",
+ "plt.title(\"The final validation loss: {}\".format(hist.history[\"val_loss\"][-1]))\n",
+ "plt.legend()\n",
+ "plt.show()\n",
"\n",
- "## Simple examples of Autoencoders"
+ "# Use the trained neural network to predict more points of the data set\n",
+ "test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])\n",
+ "# Stop the timer and calculate the total time needed.\n",
+ "end = timer()\n",
+ "print('Time: ', end-start)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Other Things to Try\n",
+ "\n",
+ "\n",
+ "Changing the size of the recurrent neural network and its parameters\n",
+ "can drastically change the results you get from the model. The below\n",
+ "code takes the simple recurrent neural network from above and adds a\n",
+ "second hidden layer, changes the number of neurons in the hidden\n",
+ "layer, and explicitly declares the activation function of the hidden\n",
+ "layers to be a sigmoid function. The loss function and optimizer can\n",
+ "also be changed but are kept the same as the above network. These\n",
+ "parameters can be tuned to provide the optimal result from the\n",
+ "network. For some ideas on how to improve the performance of a\n",
+ "[recurrent neural network](https://danijar.com/tips-for-training-recurrent-neural-networks)."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 16,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
+ "source": [
+ "def rnn_2layers(length_of_sequences, batch_size = None, stateful = False):\n",
+ " \"\"\"\n",
+ " Inputs:\n",
+ " length_of_sequences (an int): the number of y values in \"x data\". This is determined\n",
+ " when the data is formatted\n",
+ " batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.\n",
+ " stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.\n",
+ " Returns:\n",
+ " model (a Keras model): The recurrent neural network that is built and compiled by this\n",
+ " method\n",
+ " Builds and compiles a recurrent neural network with two hidden layers and returns the model.\n",
+ " \"\"\"\n",
+ " # Number of neurons in the input and output layers\n",
+ " in_out_neurons = 1\n",
+ " # Number of neurons in the hidden layer, increased from the first network\n",
+ " hidden_neurons = 500\n",
+ " # Define the input layer\n",
+ " inp = Input(batch_shape=(batch_size, \n",
+ " length_of_sequences, \n",
+ " in_out_neurons)) \n",
+ " # Create two hidden layers instead of one hidden layer. Explicitly set the activation\n",
+ " # function to be the sigmoid function (the default value is hyperbolic tangent)\n",
+ " rnn1 = SimpleRNN(hidden_neurons, \n",
+ " return_sequences=True, # This needs to be True if another hidden layer is to follow\n",
+ " stateful = stateful, activation = 'sigmoid',\n",
+ " name=\"RNN1\")(inp)\n",
+ " rnn2 = SimpleRNN(hidden_neurons, \n",
+ " return_sequences=False, activation = 'sigmoid',\n",
+ " stateful = stateful,\n",
+ " name=\"RNN2\")(rnn1)\n",
+ " # Define the output layer as a dense neural network layer (standard neural network layer)\n",
+ " #and add it to the network immediately after the hidden layer.\n",
+ " dens = Dense(in_out_neurons,name=\"dense\")(rnn2)\n",
+ " # Create the machine learning model starting with the input layer and ending with the \n",
+ " # output layer\n",
+ " model = Model(inputs=[inp],outputs=[dens])\n",
+ " # Compile the machine learning model using the mean squared error function as the loss \n",
+ " # function and an Adams optimizer.\n",
+ " model.compile(loss=\"mean_squared_error\", optimizer=\"adam\") \n",
+ " return model\n",
+ "\n",
+ "# Check to make sure the data set is complete\n",
+ "assert len(X_tot) == len(y_tot)\n",
+ "\n",
+ "# This is the number of points that will be used in as the training data\n",
+ "dim=12\n",
+ "\n",
+ "# Separate the training data from the whole data set\n",
+ "X_train = X_tot[:dim]\n",
+ "y_train = y_tot[:dim]\n",
+ "\n",
+ "\n",
+ "# Generate the training data for the RNN, using a sequence of 2\n",
+ "rnn_input, rnn_training = format_data(y_train, 2)\n",
+ "\n",
+ "\n",
+ "# Create a recurrent neural network in Keras and produce a summary of the \n",
+ "# machine learning model\n",
+ "model = rnn_2layers(length_of_sequences = 2)\n",
+ "model.summary()\n",
+ "\n",
+ "# Start the timer. Want to time training+testing\n",
+ "start = timer()\n",
+ "# Fit the model using the training data genenerated above using 150 training iterations and a 5%\n",
+ "# validation split. Setting verbose to True prints information about each training iteration.\n",
+ "hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150, \n",
+ " verbose=True,validation_split=0.05)\n",
+ "\n",
+ "\n",
+ "# This section plots the training loss and the validation loss as a function of training iteration.\n",
+ "# This is not required for analyzing the couple cluster data but can help determine if the network is\n",
+ "# being overtrained.\n",
+ "for label in [\"loss\",\"val_loss\"]:\n",
+ " plt.plot(hist.history[label],label=label)\n",
+ "\n",
+ "plt.ylabel(\"loss\")\n",
+ "plt.xlabel(\"epoch\")\n",
+ "plt.title(\"The final validation loss: {}\".format(hist.history[\"val_loss\"][-1]))\n",
+ "plt.legend()\n",
+ "plt.show()\n",
+ "\n",
+ "# Use the trained neural network to predict more points of the data set\n",
+ "test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])\n",
+ "# Stop the timer and calculate the total time needed.\n",
+ "end = timer()\n",
+ "print('Time: ', end-start)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Other Types of Recurrent Neural Networks\n",
+ "\n",
+ "Besides a simple recurrent neural network layer, there are two other\n",
+ "commonly used types of recurrent neural network layers: Long Short\n",
+ "Term Memory (LSTM) and Gated Recurrent Unit (GRU). For a short\n",
+ "introduction to these layers see \n",
+ "and .\n",
+ "\n",
+ "The first network created below is similar to the previous network,\n",
+ "but it replaces the SimpleRNN layers with LSTM layers. The second\n",
+ "network below has two hidden layers made up of GRUs, which are\n",
+ "preceeded by two dense (feeddorward) neural network layers. These\n",
+ "dense layers \"preprocess\" the data before it reaches the recurrent\n",
+ "layers. This architecture has been shown to improve the performance\n",
+ "of recurrent neural networks (see the link above and also\n",
+ "."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 17,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
+ "source": [
+ "def lstm_2layers(length_of_sequences, batch_size = None, stateful = False):\n",
+ " \"\"\"\n",
+ " Inputs:\n",
+ " length_of_sequences (an int): the number of y values in \"x data\". This is determined\n",
+ " when the data is formatted\n",
+ " batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.\n",
+ " stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.\n",
+ " Returns:\n",
+ " model (a Keras model): The recurrent neural network that is built and compiled by this\n",
+ " method\n",
+ " Builds and compiles a recurrent neural network with two LSTM hidden layers and returns the model.\n",
+ " \"\"\"\n",
+ " # Number of neurons on the input/output layer and the number of neurons in the hidden layer\n",
+ " in_out_neurons = 1\n",
+ " hidden_neurons = 250\n",
+ " # Input Layer\n",
+ " inp = Input(batch_shape=(batch_size, \n",
+ " length_of_sequences, \n",
+ " in_out_neurons)) \n",
+ " # Hidden layers (in this case they are LSTM layers instead if SimpleRNN layers)\n",
+ " rnn= LSTM(hidden_neurons, \n",
+ " return_sequences=True,\n",
+ " stateful = stateful,\n",
+ " name=\"RNN\", use_bias=True, activation='tanh')(inp)\n",
+ " rnn1 = LSTM(hidden_neurons, \n",
+ " return_sequences=False,\n",
+ " stateful = stateful,\n",
+ " name=\"RNN1\", use_bias=True, activation='tanh')(rnn)\n",
+ " # Output layer\n",
+ " dens = Dense(in_out_neurons,name=\"dense\")(rnn1)\n",
+ " # Define the midel\n",
+ " model = Model(inputs=[inp],outputs=[dens])\n",
+ " # Compile the model\n",
+ " model.compile(loss='mean_squared_error', optimizer='adam') \n",
+ " # Return the model\n",
+ " return model\n",
+ "\n",
+ "def dnn2_gru2(length_of_sequences, batch_size = None, stateful = False):\n",
+ " \"\"\"\n",
+ " Inputs:\n",
+ " length_of_sequences (an int): the number of y values in \"x data\". This is determined\n",
+ " when the data is formatted\n",
+ " batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.\n",
+ " stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.\n",
+ " Returns:\n",
+ " model (a Keras model): The recurrent neural network that is built and compiled by this\n",
+ " method\n",
+ " Builds and compiles a recurrent neural network with four hidden layers (two dense followed by\n",
+ " two GRU layers) and returns the model.\n",
+ " \"\"\" \n",
+ " # Number of neurons on the input/output layers and hidden layers\n",
+ " in_out_neurons = 1\n",
+ " hidden_neurons = 250\n",
+ " # Input layer\n",
+ " inp = Input(batch_shape=(batch_size, \n",
+ " length_of_sequences, \n",
+ " in_out_neurons)) \n",
+ " # Hidden Dense (feedforward) layers\n",
+ " dnn = Dense(hidden_neurons/2, activation='relu', name='dnn')(inp)\n",
+ " dnn1 = Dense(hidden_neurons/2, activation='relu', name='dnn1')(dnn)\n",
+ " # Hidden GRU layers\n",
+ " rnn1 = GRU(hidden_neurons, \n",
+ " return_sequences=True,\n",
+ " stateful = stateful,\n",
+ " name=\"RNN1\", use_bias=True)(dnn1)\n",
+ " rnn = GRU(hidden_neurons, \n",
+ " return_sequences=False,\n",
+ " stateful = stateful,\n",
+ " name=\"RNN\", use_bias=True)(rnn1)\n",
+ " # Output layer\n",
+ " dens = Dense(in_out_neurons,name=\"dense\")(rnn)\n",
+ " # Define the model\n",
+ " model = Model(inputs=[inp],outputs=[dens])\n",
+ " # Compile the mdoel\n",
+ " model.compile(loss='mean_squared_error', optimizer='adam') \n",
+ " # Return the model\n",
+ " return model\n",
+ "\n",
+ "# Check to make sure the data set is complete\n",
+ "assert len(X_tot) == len(y_tot)\n",
+ "\n",
+ "# This is the number of points that will be used in as the training data\n",
+ "dim=12\n",
+ "\n",
+ "# Separate the training data from the whole data set\n",
+ "X_train = X_tot[:dim]\n",
+ "y_train = y_tot[:dim]\n",
+ "\n",
+ "\n",
+ "# Generate the training data for the RNN, using a sequence of 2\n",
+ "rnn_input, rnn_training = format_data(y_train, 2)\n",
+ "\n",
+ "\n",
+ "# Create a recurrent neural network in Keras and produce a summary of the \n",
+ "# machine learning model\n",
+ "# Change the method name to reflect which network you want to use\n",
+ "model = dnn2_gru2(length_of_sequences = 2)\n",
+ "model.summary()\n",
+ "\n",
+ "# Start the timer. Want to time training+testing\n",
+ "start = timer()\n",
+ "# Fit the model using the training data genenerated above using 150 training iterations and a 5%\n",
+ "# validation split. Setting verbose to True prints information about each training iteration.\n",
+ "hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150, \n",
+ " verbose=True,validation_split=0.05)\n",
+ "\n",
+ "\n",
+ "# This section plots the training loss and the validation loss as a function of training iteration.\n",
+ "# This is not required for analyzing the couple cluster data but can help determine if the network is\n",
+ "# being overtrained.\n",
+ "for label in [\"loss\",\"val_loss\"]:\n",
+ " plt.plot(hist.history[label],label=label)\n",
+ "\n",
+ "plt.ylabel(\"loss\")\n",
+ "plt.xlabel(\"epoch\")\n",
+ "plt.title(\"The final validation loss: {}\".format(hist.history[\"val_loss\"][-1]))\n",
+ "plt.legend()\n",
+ "plt.show()\n",
+ "\n",
+ "# Use the trained neural network to predict more points of the data set\n",
+ "test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])\n",
+ "# Stop the timer and calculate the total time needed.\n",
+ "end = timer()\n",
+ "print('Time: ', end-start)\n",
+ "\n",
+ "\n",
+ "# ### Training Recurrent Neural Networks in the Standard Way (i.e. learning the relationship between the X and Y data)\n",
+ "# \n",
+ "# Finally, comparing the performace of a recurrent neural network using the standard data formatting to the performance of the network with time sequence data formatting shows the benefit of this type of data formatting with extrapolation.\n",
+ "\n",
+ "# Check to make sure the data set is complete\n",
+ "assert len(X_tot) == len(y_tot)\n",
+ "\n",
+ "# This is the number of points that will be used in as the training data\n",
+ "dim=12\n",
+ "\n",
+ "# Separate the training data from the whole data set\n",
+ "X_train = X_tot[:dim]\n",
+ "y_train = y_tot[:dim]\n",
+ "\n",
+ "# Reshape the data for Keras specifications\n",
+ "X_train = X_train.reshape((dim, 1))\n",
+ "y_train = y_train.reshape((dim, 1))\n",
+ "\n",
+ "\n",
+ "# Create a recurrent neural network in Keras and produce a summary of the \n",
+ "# machine learning model\n",
+ "# Set the sequence length to 1 for regular data formatting \n",
+ "model = rnn(length_of_sequences = 1)\n",
+ "model.summary()\n",
+ "\n",
+ "# Start the timer. Want to time training+testing\n",
+ "start = timer()\n",
+ "# Fit the model using the training data genenerated above using 150 training iterations and a 5%\n",
+ "# validation split. Setting verbose to True prints information about each training iteration.\n",
+ "hist = model.fit(X_train, y_train, batch_size=None, epochs=150, \n",
+ " verbose=True,validation_split=0.05)\n",
+ "\n",
+ "\n",
+ "# This section plots the training loss and the validation loss as a function of training iteration.\n",
+ "# This is not required for analyzing the couple cluster data but can help determine if the network is\n",
+ "# being overtrained.\n",
+ "for label in [\"loss\",\"val_loss\"]:\n",
+ " plt.plot(hist.history[label],label=label)\n",
+ "\n",
+ "plt.ylabel(\"loss\")\n",
+ "plt.xlabel(\"epoch\")\n",
+ "plt.title(\"The final validation loss: {}\".format(hist.history[\"val_loss\"][-1]))\n",
+ "plt.legend()\n",
+ "plt.show()\n",
+ "\n",
+ "# Use the trained neural network to predict the remaining data points\n",
+ "X_pred = X_tot[dim:]\n",
+ "X_pred = X_pred.reshape((len(X_pred), 1))\n",
+ "y_model = model.predict(X_pred)\n",
+ "y_pred = np.concatenate((y_tot[:dim], y_model.flatten()))\n",
+ "\n",
+ "# Plot the known data set and the predicted data set. The red box represents the region that was used\n",
+ "# for the training data.\n",
+ "fig, ax = plt.subplots()\n",
+ "ax.plot(X_tot, y_tot, label=\"true\", linewidth=3)\n",
+ "ax.plot(X_tot, y_pred, 'g-.',label=\"predicted\", linewidth=4)\n",
+ "ax.legend()\n",
+ "# Created a red region to represent the points used in the training data.\n",
+ "ax.axvspan(X_tot[0], X_tot[dim], alpha=0.25, color='red')\n",
+ "plt.show()\n",
+ "\n",
+ "# Stop the timer and calculate the total time needed.\n",
+ "end = timer()\n",
+ "print('Time: ', end-start)"
]
}
],
- "metadata": {
- "kernelspec": {
- "display_name": "Python 3",
- "language": "python",
- "name": "python3"
- },
- "language_info": {
- "codemirror_mode": {
- "name": "ipython",
- "version": 3
- },
- "file_extension": ".py",
- "mimetype": "text/x-python",
- "name": "python",
- "nbconvert_exporter": "python",
- "pygments_lexer": "ipython3",
- "version": "3.6.8"
- }
- },
+ "metadata": {},
"nbformat": 4,
"nbformat_minor": 4
}
diff --git a/doc/src/week42/week42.do.txt b/doc/src/week42/week42.do.txt
index 3e2501ba4..2db6df215 100644
--- a/doc/src/week42/week42.do.txt
+++ b/doc/src/week42/week42.do.txt
@@ -1,4 +1,4 @@
-TITLE: Week 42 Convolutional and Recurrent Neural Networks and Autoencoders
+TITLE: Week 42 Convolutional (CNN) and Recurrent (RNN) Neural Networks
AUTHOR: Morten Hjorth-Jensen {copyright, 1999-present|CC BY-NC} at Department of Physics, University of Oslo & Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University
DATE: today
@@ -6,11 +6,16 @@ DATE: today
!split
===== Plan for week 42 =====
-* Thursday: Convolutional Neural Networks and examples
-* Friday: Recurrent Neural Networks and Autoencoders
+* Thursday: Convolutional Neural Networks and examples. "Video of Lecture":"https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureOctober15.mp4?vrtx=view-as-webpage"
+* Friday: Recurrent Neural Networks. "Video of Lecture":"https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureOctober16.mp4?vrtx=view-as-webpage"
Reading suggestions for both days: "Aurelien Geron's chapters 13 and 14":"https://github.com/CompPhysics/MachineLearning/blob/master/doc/Textbooks/TensorflowML.pdf". Autoencoders are discussed in chapter 15 of Geron's text.
+!bblock Excellent lectures on CNNs and RNNs
+* "Video on Convolutional Neural Networks from MIT":"https://www.youtube.com/watch?v=iaSUYvmCekI&ab_channel=AlexanderAmini"
+* "Video on Recurrent Neural Networks from MIT":"https://www.youtube.com/watch?v=SEnXr6v2ifU&ab_channel=AlexanderAmini"
+!eblock
+
@@ -589,6 +594,16 @@ input, making them extremely useful for natural language processing
systems such as automatic translation and speech-to-text.
+
+
+!split
+===== Set up of an RNN =====
+
+See the "handwritten notes":"https://github.com/CompPhysics/MachineLearning/blob/master/doc/HandWrittenNotes/NotesOctober16.pdf" and the "video from the lecture of October 16":"https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureOctober16.mp4?vrtx=view-as-webpage".
+
+More text will be added later.
+
+
!split
===== A simple example =====
@@ -665,76 +680,570 @@ plt.show()
!split
-===== Set up of an RNN =====
+===== An extrapolation example =====
-The figure here displays a simple example of an RNN, with inputs $x_t$
-at a given time $t$ and outputs $y_t$. Introducing time as a variable
-offers an intutitive way of understanding these networks. In addition
-to the inputs $x_t$, the layer at a time $t$ receives also as input
-the output from the previous layer $t-1$, that is $y_{t1}$.
+The following code provides an example of how recurrent neural
+networks can be used to extrapolate to unknown values of physics data
+sets. Specifically, the data sets used in this program come from
+a quantum mechanical many-body calculation of energies as functions of the number of particles.
-This means also that we need to have weights that link both the inputs
-$x_t$ to the outputs $y_t$ as well as weights that link the output
-from the previous time $y_{t-1}$ and $y_t$. The figure here shows an
-example of a simple RNN.
-More material will be added here.
+!bc pycod
+# For matrices and calculations
+import numpy as np
+# For machine learning (backend for keras)
+import tensorflow as tf
+# User-friendly machine learning library
+# Front end for TensorFlow
+import tensorflow.keras
+# Different methods from Keras needed to create an RNN
+# This is not necessary but it shortened function calls
+# that need to be used in the code.
+from tensorflow.keras import datasets, layers, models
+from tensorflow.keras.layers import Input
+from tensorflow.keras import regularizers
+from tensorflow.keras.models import Model, Sequential
+from tensorflow.keras.layers import Dense, SimpleRNN, LSTM, GRU
+# For timing the code
+from timeit import default_timer as timer
+# For plotting
+import matplotlib.pyplot as plt
+
+
+# The data set
+datatype='VaryDimension'
+X_tot = np.arange(2, 42, 2)
+y_tot = np.array([-0.03077640549, -0.08336233266, -0.1446729567, -0.2116753732, -0.2830637392, -0.3581341341, -0.436462435, -0.5177783846,
+ -0.6019067271, -0.6887363571, -0.7782028952, -0.8702784034, -0.9649652536, -1.062292565, -1.16231451,
+ -1.265109911, -1.370782966, -1.479465113, -1.591317992, -1.70653767])
+
+!ec
!split
-===== Solving differential equations and eigenvalue problems with RNNs =====
+===== Formatting the Data =====
+
+The way the recurrent neural networks are trained in this program
+differs from how machine learning algorithms are usually trained.
+Typically a machine learning algorithm is trained by learning the
+relationship between the x data and the y data. In this program, the
+recurrent neural network will be trained to recognize the relationship
+in a sequence of y values. This is type of data formatting is
+typically used time series forcasting, but it can also be used in any
+extrapolation (time series forecasting is just a specific type of
+extrapolation along the time axis). This method of data formatting
+does not use the x data and assumes that the y data are evenly spaced.
+
+For a standard machine learning algorithm, the training data has the
+form of (x,y) so the machine learning algorithm learns to assiciate a
+y value with a given x value. This is useful when the test data has x
+values within the same range as the training data. However, for this
+application, the x values of the test data are outside of the x values
+of the training data and the traditional method of training a machine
+learning algorithm does not work as well. For this reason, the
+recurrent neural network is trained on sequences of y values of the
+form ((y1, y2), y3), so that the network is concerned with learning
+the pattern of the y data and not the relation between the x and y
+data. As long as the pattern of y data outside of the training region
+stays relatively stable compared to what was inside the training
+region, this method of training can produce accurate extrapolations to
+y values far removed from the training data set.
+#
+# The idea behind formatting the data in this way comes from [this resource](https://machinelearningmastery.com/time-series-prediction-lstm-recurrent-neural-networks-python-keras/) and [this one](https://fairyonice.github.io/Understand-Keras%27s-RNN-behind-the-scenes-with-a-sin-wave-example.html).
+#
+# The following method takes in a y data set and formats it so the "x data" are of the form (y1, y2) and the "y data" are of the form y3, with extra brackets added in to make the resulting arrays compatable with both Keras and Tensorflow.
+#
+# Note: Using a sequence length of two is not required for time series forecasting so any lenght of sequence could be used (for example instead of ((y1, y2) y3) you could change the length of sequence to be 4 and the resulting data points would have the form ((y1, y2, y3, y4), y5)). While the following method can be used to create a data set of any sequence length, the remainder of the code expects the length of sequence to be 2. This is because the data sets are very small and the higher the lenght of the sequence the less resulting data points.
-In our discussions of ordinary differential equations and partial
-differential equations using neural networks. Here we will discuss how
-we can solve say ordinary differential equations and eigenvalue
-problems using RNNs. Eigenvalue problems can be solved using RNNs by
-rewriting such a problems as a non-linear differential equation.
+!bc pycod
+# FORMAT_DATA
+def format_data(data, length_of_sequence = 2):
+ """
+ Inputs:
+ data(a numpy array): the data that will be the inputs to the recurrent neural
+ network
+ length_of_sequence (an int): the number of elements in one iteration of the
+ sequence patter. For a function approximator use length_of_sequence = 2.
+ Returns:
+ rnn_input (a 3D numpy array): the input data for the recurrent neural network. Its
+ dimensions are length of data - length of sequence, length of sequence,
+ dimnsion of data
+ rnn_output (a numpy array): the training data for the neural network
+ Formats data to be used in a recurrent neural network.
+ """
-Instead of starting with a well-known ordinary differential equation,
-we start directly with an eigenvaule problem.
+ X, Y = [], []
+ for i in range(len(data)-length_of_sequence):
+ # Get the next length_of_sequence elements
+ a = data[i:i+length_of_sequence]
+ # Get the element that immediately follows that
+ b = data[i+length_of_sequence]
+ # Reshape so that each data point is contained in its own array
+ a = np.reshape (a, (len(a), 1))
+ X.append(a)
+ Y.append(b)
+ rnn_input = np.array(X)
+ rnn_output = np.array(Y)
+
+ return rnn_input, rnn_output
+# ## Defining the Recurrent Neural Network Using Keras
+#
+# The following method defines a simple recurrent neural network in keras consisting of one input layer, one hidden layer, and one output layer.
+
+def rnn(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with one hidden layer and returns the model.
+ """
+ # Number of neurons in the input and output layers
+ in_out_neurons = 1
+ # Number of neurons in the hidden layer
+ hidden_neurons = 200
+ # Define the input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Define the hidden layer as a simple RNN layer with a set number of neurons and add it to
+ # the network immediately after the input layer
+ rnn = SimpleRNN(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN")(inp)
+ # Define the output layer as a dense neural network layer (standard neural network layer)
+ #and add it to the network immediately after the hidden layer.
+ dens = Dense(in_out_neurons,name="dense")(rnn)
+ # Create the machine learning model starting with the input layer and ending with the
+ # output layer
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the machine learning model using the mean squared error function as the loss
+ # function and an Adams optimizer.
+ model.compile(loss="mean_squared_error", optimizer="adam")
+ return model
+
+!ec
!split
-===== Long-Short Time Memory =====
+===== Predicting New Points With A Trained Recurrent Neural Network =====
-Discussions about dynamic unrolling through time. discuss memory cells, input and output
+!bc pycod
+def test_rnn (x1, y_test, plot_min, plot_max):
+ """
+ Inputs:
+ x1 (a list or numpy array): The complete x component of the data set
+ y_test (a list or numpy array): The complete y component of the data set
+ plot_min (an int or float): the smallest x value used in the training data
+ plot_max (an int or float): the largest x valye used in the training data
+ Returns:
+ None.
+ Uses a trained recurrent neural network model to predict future points in the
+ series. Computes the MSE of the predicted data set from the true data set, saves
+ the predicted data set to a csv file, and plots the predicted and true data sets w
+ while also displaying the data range used for training.
+ """
+ # Add the training data as the first dim points in the predicted data array as these
+ # are known values.
+ y_pred = y_test[:dim].tolist()
+ # Generate the first input to the trained recurrent neural network using the last two
+ # points of the training data. Based on how the network was trained this means that it
+ # will predict the first point in the data set after the training data. All of the
+ # brackets are necessary for Tensorflow.
+ next_input = np.array([[[y_test[dim-2]], [y_test[dim-1]]]])
+ # Save the very last point in the training data set. This will be used later.
+ last = [y_test[dim-1]]
+
+ # Iterate until the complete data set is created.
+ for i in range (dim, len(y_test)):
+ # Predict the next point in the data set using the previous two points.
+ next = model.predict(next_input)
+ # Append just the number of the predicted data set
+ y_pred.append(next[0][0])
+ # Create the input that will be used to predict the next data point in the data set.
+ next_input = np.array([[last, next[0]]], dtype=np.float64)
+ last = next
+
+ # Print the mean squared error between the known data set and the predicted data set.
+ print('MSE: ', np.square(np.subtract(y_test, y_pred)).mean())
+ # Save the predicted data set as a csv file for later use
+ name = datatype + 'Predicted'+str(dim)+'.csv'
+ np.savetxt(name, y_pred, delimiter=',')
+ # Plot the known data set and the predicted data set. The red box represents the region that was used
+ # for the training data.
+ fig, ax = plt.subplots()
+ ax.plot(x1, y_test, label="true", linewidth=3)
+ ax.plot(x1, y_pred, 'g-.',label="predicted", linewidth=4)
+ ax.legend()
+ # Created a red region to represent the points used in the training data.
+ ax.axvspan(plot_min, plot_max, alpha=0.25, color='red')
+ plt.show()
+
+# Check to make sure the data set is complete
+assert len(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+model = rnn(length_of_sequences = rnn_input.shape[1])
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+!ec
+
!split
-===== Autoencoders: Overarching view =====
+===== Other Things to Try =====
-Autoencoders are artificial neural networks capable of learning
-efficient representations of the input data (these representations are called codings) without
-any supervision (i.e., the training set is unlabeled). These codings
-typically have a much lower dimensionality than the input data, making
-autoencoders useful for dimensionality reduction.
-More importantly, autoencoders act as powerful feature detectors, and
-they can be used for unsupervised pretraining of deep neural networks.
+Changing the size of the recurrent neural network and its parameters
+can drastically change the results you get from the model. The below
+code takes the simple recurrent neural network from above and adds a
+second hidden layer, changes the number of neurons in the hidden
+layer, and explicitly declares the activation function of the hidden
+layers to be a sigmoid function. The loss function and optimizer can
+also be changed but are kept the same as the above network. These
+parameters can be tuned to provide the optimal result from the
+network. For some ideas on how to improve the performance of a
+"recurrent neural network":"https://danijar.com/tips-for-training-recurrent-neural-networks".
-Lastly, they are capable of randomly generating new data that looks
-very similar to the training data; this is called a generative
-model. For example, you could train an autoencoder on pictures of
-faces, and it would then be able to generate new faces. Surprisingly,
-autoencoders work by simply learning to copy their inputs to their
-outputs. This may sound like a trivial task, but we will see that
-constraining the network in various ways can make it rather
-difficult. For example, you can limit the size of the internal
-representation, or you can add noise to the inputs and train the
-network to recover the original inputs. These constraints prevent the
-autoencoder from trivially copying the inputs directly to the outputs,
-which forces it to learn efficient ways of representing the data. In
-short, the codings are byproducts of the autoencoder’s attempt to
-learn the identity function under some constraints.
+!bc pycod
+def rnn_2layers(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with two hidden layers and returns the model.
+ """
+ # Number of neurons in the input and output layers
+ in_out_neurons = 1
+ # Number of neurons in the hidden layer, increased from the first network
+ hidden_neurons = 500
+ # Define the input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Create two hidden layers instead of one hidden layer. Explicitly set the activation
+ # function to be the sigmoid function (the default value is hyperbolic tangent)
+ rnn1 = SimpleRNN(hidden_neurons,
+ return_sequences=True, # This needs to be True if another hidden layer is to follow
+ stateful = stateful, activation = 'sigmoid',
+ name="RNN1")(inp)
+ rnn2 = SimpleRNN(hidden_neurons,
+ return_sequences=False, activation = 'sigmoid',
+ stateful = stateful,
+ name="RNN2")(rnn1)
+ # Define the output layer as a dense neural network layer (standard neural network layer)
+ #and add it to the network immediately after the hidden layer.
+ dens = Dense(in_out_neurons,name="dense")(rnn2)
+ # Create the machine learning model starting with the input layer and ending with the
+ # output layer
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the machine learning model using the mean squared error function as the loss
+ # function and an Adams optimizer.
+ model.compile(loss="mean_squared_error", optimizer="adam")
+ return model
+
+# Check to make sure the data set is complete
+assert len(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+model = rnn_2layers(length_of_sequences = 2)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+!ec
!split
-===== Simple examples of Autoencoders =====
+===== Other Types of Recurrent Neural Networks =====
+
+Besides a simple recurrent neural network layer, there are two other
+commonly used types of recurrent neural network layers: Long Short
+Term Memory (LSTM) and Gated Recurrent Unit (GRU). For a short
+introduction to these layers see URL:"https://medium.com/mindboard/lstm-vs-gru-experimental-comparison-955820c21e8b"
+and URL:"https://medium.com/mindboard/lstm-vs-gru-experimental-comparison-955820c21e8b".
+
+The first network created below is similar to the previous network,
+but it replaces the SimpleRNN layers with LSTM layers. The second
+network below has two hidden layers made up of GRUs, which are
+preceeded by two dense (feeddorward) neural network layers. These
+dense layers "preprocess" the data before it reaches the recurrent
+layers. This architecture has been shown to improve the performance
+of recurrent neural networks (see the link above and also
+URL:"https://arxiv.org/pdf/1807.02857.pdf".
+
+!bc pycod
+def lstm_2layers(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with two LSTM hidden layers and returns the model.
+ """
+ # Number of neurons on the input/output layer and the number of neurons in the hidden layer
+ in_out_neurons = 1
+ hidden_neurons = 250
+ # Input Layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Hidden layers (in this case they are LSTM layers instead if SimpleRNN layers)
+ rnn= LSTM(hidden_neurons,
+ return_sequences=True,
+ stateful = stateful,
+ name="RNN", use_bias=True, activation='tanh')(inp)
+ rnn1 = LSTM(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN1", use_bias=True, activation='tanh')(rnn)
+ # Output layer
+ dens = Dense(in_out_neurons,name="dense")(rnn1)
+ # Define the midel
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the model
+ model.compile(loss='mean_squared_error', optimizer='adam')
+ # Return the model
+ return model
+
+def dnn2_gru2(length_of_sequences, batch_size = None, stateful = False):
+ """
+ Inputs:
+ length_of_sequences (an int): the number of y values in "x data". This is determined
+ when the data is formatted
+ batch_size (an int): Default value is None. See Keras documentation of SimpleRNN.
+ stateful (a boolean): Default value is False. See Keras documentation of SimpleRNN.
+ Returns:
+ model (a Keras model): The recurrent neural network that is built and compiled by this
+ method
+ Builds and compiles a recurrent neural network with four hidden layers (two dense followed by
+ two GRU layers) and returns the model.
+ """
+ # Number of neurons on the input/output layers and hidden layers
+ in_out_neurons = 1
+ hidden_neurons = 250
+ # Input layer
+ inp = Input(batch_shape=(batch_size,
+ length_of_sequences,
+ in_out_neurons))
+ # Hidden Dense (feedforward) layers
+ dnn = Dense(hidden_neurons/2, activation='relu', name='dnn')(inp)
+ dnn1 = Dense(hidden_neurons/2, activation='relu', name='dnn1')(dnn)
+ # Hidden GRU layers
+ rnn1 = GRU(hidden_neurons,
+ return_sequences=True,
+ stateful = stateful,
+ name="RNN1", use_bias=True)(dnn1)
+ rnn = GRU(hidden_neurons,
+ return_sequences=False,
+ stateful = stateful,
+ name="RNN", use_bias=True)(rnn1)
+ # Output layer
+ dens = Dense(in_out_neurons,name="dense")(rnn)
+ # Define the model
+ model = Model(inputs=[inp],outputs=[dens])
+ # Compile the mdoel
+ model.compile(loss='mean_squared_error', optimizer='adam')
+ # Return the model
+ return model
+
+# Check to make sure the data set is complete
+assert len(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+# Generate the training data for the RNN, using a sequence of 2
+rnn_input, rnn_training = format_data(y_train, 2)
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+# Change the method name to reflect which network you want to use
+model = dnn2_gru2(length_of_sequences = 2)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(rnn_input, rnn_training, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict more points of the data set
+test_rnn(X_tot, y_tot, X_tot[0], X_tot[dim-1])
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
+
+# ### Training Recurrent Neural Networks in the Standard Way (i.e. learning the relationship between the X and Y data)
+#
+# Finally, comparing the performace of a recurrent neural network using the standard data formatting to the performance of the network with time sequence data formatting shows the benefit of this type of data formatting with extrapolation.
+
+# Check to make sure the data set is complete
+assert len(X_tot) == len(y_tot)
+
+# This is the number of points that will be used in as the training data
+dim=12
+
+# Separate the training data from the whole data set
+X_train = X_tot[:dim]
+y_train = y_tot[:dim]
+
+# Reshape the data for Keras specifications
+X_train = X_train.reshape((dim, 1))
+y_train = y_train.reshape((dim, 1))
+
+
+# Create a recurrent neural network in Keras and produce a summary of the
+# machine learning model
+# Set the sequence length to 1 for regular data formatting
+model = rnn(length_of_sequences = 1)
+model.summary()
+
+# Start the timer. Want to time training+testing
+start = timer()
+# Fit the model using the training data genenerated above using 150 training iterations and a 5%
+# validation split. Setting verbose to True prints information about each training iteration.
+hist = model.fit(X_train, y_train, batch_size=None, epochs=150,
+ verbose=True,validation_split=0.05)
+
+
+# This section plots the training loss and the validation loss as a function of training iteration.
+# This is not required for analyzing the couple cluster data but can help determine if the network is
+# being overtrained.
+for label in ["loss","val_loss"]:
+ plt.plot(hist.history[label],label=label)
+
+plt.ylabel("loss")
+plt.xlabel("epoch")
+plt.title("The final validation loss: {}".format(hist.history["val_loss"][-1]))
+plt.legend()
+plt.show()
+
+# Use the trained neural network to predict the remaining data points
+X_pred = X_tot[dim:]
+X_pred = X_pred.reshape((len(X_pred), 1))
+y_model = model.predict(X_pred)
+y_pred = np.concatenate((y_tot[:dim], y_model.flatten()))
+
+# Plot the known data set and the predicted data set. The red box represents the region that was used
+# for the training data.
+fig, ax = plt.subplots()
+ax.plot(X_tot, y_tot, label="true", linewidth=3)
+ax.plot(X_tot, y_pred, 'g-.',label="predicted", linewidth=4)
+ax.legend()
+# Created a red region to represent the points used in the training data.
+ax.axvspan(X_tot[0], X_tot[dim], alpha=0.25, color='red')
+plt.show()
+
+# Stop the timer and calculate the total time needed.
+end = timer()
+print('Time: ', end-start)
+
+!ec