diff --git a/doc/pub/DimRed/html/._DimRed-bs000.html b/doc/pub/DimRed/html/._DimRed-bs000.html index 7f403ef8e..b87451a93 100644 --- a/doc/pub/DimRed/html/._DimRed-bs000.html +++ b/doc/pub/DimRed/html/._DimRed-bs000.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -160,7 +174,7 @@ MathJax.Hub.Config({
    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

    -

    Oct 17, 2019

    +

    Oct 19, 2019


    @@ -184,7 +198,7 @@ MathJax.Hub.Config({

  • 9
  • 10
  • ...
    • -
  • 17
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs001.html b/doc/pub/DimRed/html/._DimRed-bs001.html index 5e0d8bd11..07788dd92 100644 --- a/doc/pub/DimRed/html/._DimRed-bs001.html +++ b/doc/pub/DimRed/html/._DimRed-bs001.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -147,16 +161,22 @@ MathJax.Hub.Config({

    -Many Machine Learning problems involve thousands or even millions of features for each training -instance. Not only does this make training extremely slow, it can also make it much harder to find a good -solution, as we will see. This problem is often referred to as the curse of dimensionality. -Fortunately, in real-world problems, it is often possible to reduce the number of features considerably, -turning an intractable problem into a tractable one. +Many Machine Learning problems involve thousands or even millions of +features for each training instance. Not only does this make training +extremely slow, it can also make it much harder to find a good +solution, as we will see. This problem is often referred to as the +curse of dimensionality. Fortunately, in real-world problems, it is +often possible to reduce the number of features considerably, turning +an intractable problem into a tractable one. + +

    +Here we will discuss some of the most popular dimensionality reduction +techniques: the principal component analysis PCA, Kernel PCA, and +Locally Linear Embedding (LLE). Furthermore, we will start by looking +at some simple preprocessing of the data which allow us to rescale the +data.

    -Here we will discuss some of the most popular dimensionality -reduction techniques: the principal component analysis PCA, Kernel PCA, and Locally Linear Embedding (LLE). -Furthermore, we will start by looking at some simple preprocessing of the data which allow us to rescale the data. @@ -178,7 +198,7 @@ Furthermore, we will start by looking at some simple preprocessing of the data w

  • 10
  • 11
  • ...
    • -
  • 17
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs002.html b/doc/pub/DimRed/html/._DimRed-bs002.html index dcad38aec..40e9fbfa3 100644 --- a/doc/pub/DimRed/html/._DimRed-bs002.html +++ b/doc/pub/DimRed/html/._DimRed-bs002.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -148,11 +162,11 @@ MathJax.Hub.Config({

    Before we proceed however, we will discuss how to preprocess our -data. Till now and in connection with our previous examples we have not met so many cases -where we are too sensitive to the scaling of our data. Normally the -data may need a rescaling and/or may be sensitive to extreme -values. Scaling the data renders our inputs much more suitable for the -algorithms we want to employ. +data. Till now and in connection with our previous examples we have +not met so many cases where we are too sensitive to the scaling of our +data. Normally the data may need a rescaling and/or may be sensitive +to extreme values. Scaling the data renders our inputs much more +suitable for the algorithms we want to employ.

    Scikit-Learn has several functions which allow us to rescale the @@ -183,7 +197,7 @@ ensures that all features are exactly between \( 0 \) and \( 1 \). The

  • 11
  • 12
  • ...
    • -
  • 17
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs003.html b/doc/pub/DimRed/html/._DimRed-bs003.html index a60425757..1ec740a66 100644 --- a/doc/pub/DimRed/html/._DimRed-bs003.html +++ b/doc/pub/DimRed/html/._DimRed-bs003.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -186,7 +200,7 @@ techniques.
  • 12
  • 13
  • ...
    • -
  • 17
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs004.html b/doc/pub/DimRed/html/._DimRed-bs004.html index c049c3a2c..d5e0a2e1f 100644 --- a/doc/pub/DimRed/html/._DimRed-bs004.html +++ b/doc/pub/DimRed/html/._DimRed-bs004.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -261,7 +275,7 @@ svm.fit(X_train_scaled, y_train)
  • 13
  • 14
  • ...
    • -
  • 17
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs005.html b/doc/pub/DimRed/html/._DimRed-bs005.html index f6f0c9b95..b9d0cd7d3 100644 --- a/doc/pub/DimRed/html/._DimRed-bs005.html +++ b/doc/pub/DimRed/html/._DimRed-bs005.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -211,7 +225,7 @@ svm.fit(X_train_scaled, y_train)
  • 14
  • 15
  • ...
    • -
  • 17
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs006.html b/doc/pub/DimRed/html/._DimRed-bs006.html index 3db112a18..60eec4c18 100644 --- a/doc/pub/DimRed/html/._DimRed-bs006.html +++ b/doc/pub/DimRed/html/._DimRed-bs006.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -190,7 +204,7 @@ logreg.fit(X_train_scaled, y_train)
  • 15
  • 16
  • ...
    • -
  • 17
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs007.html b/doc/pub/DimRed/html/._DimRed-bs007.html index 0334cf680..ce8c904ff 100644 --- a/doc/pub/DimRed/html/._DimRed-bs007.html +++ b/doc/pub/DimRed/html/._DimRed-bs007.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -205,6 +219,9 @@ X_test_scaled = scaler.fit(X_train_scaled, y_train) print("Test set accuracy scaled data: {:.2f}".format(logreg.score(X_test_scaled,y_test))) +

    + +

    @@ -227,6 +244,8 @@ logreg.fit(X_train_scaled, y_train)

  • 15
  • 16
  • 17
    • +
  • ...
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs008.html b/doc/pub/DimRed/html/._DimRed-bs008.html index c2c2bd543..61e98700f 100644 --- a/doc/pub/DimRed/html/._DimRed-bs008.html +++ b/doc/pub/DimRed/html/._DimRed-bs008.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -141,20 +155,8 @@ MathJax.Hub.Config({ -

    Getting started with PCA

    +

    Basic ideas of the Principal Component Analysis (PCA)

    -

    -This material is being finalized, not yet ready. -

    - - -

    # Now add PCA
-from sklearn.decomposition import PCA
-pca = PCA(n_components = 2)
-pca.fit(X_train_scaled)
-
-X_pca = pca.transform(X_train_scaled)
-

    @@ -177,6 +179,9 @@ X_pca = pca.15

  • 16
  • 17
    • +
  • 18
    • +
  • ...
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs009.html b/doc/pub/DimRed/html/._DimRed-bs009.html index 2249e86a0..8481e8279 100644 --- a/doc/pub/DimRed/html/._DimRed-bs009.html +++ b/doc/pub/DimRed/html/._DimRed-bs009.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -141,39 +155,8 @@ MathJax.Hub.Config({ -

    Principal Component Analysis

    -
    -
    -

    -Principal Component Analysis (PCA) is by far the most popular dimensionality reduction algorithm. -First it identifies the hyperplane that lies closest to the data, and then it projects the data onto it. +

    Introducing the Covariance and Correlation functions

    -

    -The following Python code uses NumPy’s svd() function to obtain all the principal components of the -training set, then extracts the first two principal components -

    - - -

    X_centered = X - X.mean(axis=0)
-U, s, V = np.linalg.svd(X_centered)
-c1 = V.T[:, 0]
-c2 = V.T[:, 1]
-
    -

    -PCA assumes that the dataset is centered around the origin. Scikit-Learn’s PCA classes take care of centering -the data for you. However, if you implement PCA yourself (as in the preceding example), or if you use other libraries, don’t -forget to center the data first. - -

    -Once you have identified all the principal components, you can reduce the dimensionality of the dataset -down to \( d \) dimensions by projecting it onto the hyperplane defined by the first \( d \) principal components. -Selecting this hyperplane ensures that the projection will preserve as much variance as possible. -

    - - -

    W2 = V.T[:, :2]
-X2D = X_centered.dot(W2)
-

    @@ -196,6 +179,10 @@ X2D = X_centered15

  • 16
  • 17
    • +
  • 18
    • +
  • 19
    • +
  • ...
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/DimRed-bs.html b/doc/pub/DimRed/html/DimRed-bs.html index 7f403ef8e..b87451a93 100644 --- a/doc/pub/DimRed/html/DimRed-bs.html +++ b/doc/pub/DimRed/html/DimRed-bs.html @@ -64,15 +64,25 @@ Automatically generated HTML file from DocOnce source 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -117,15 +127,19 @@ MathJax.Hub.Config({
  • Simple preprocessing examples, breast cancer data and classification, Support Vector Machines
  • More on Cancer Data, now with Logistic Regression
  • Why should we think of reducing the dimensionality
    • -
  • Getting started with PCA
    • -
  • Principal Component Analysis
    • -
  • PCA and scikit-learn
    • -
  • More on the PCA
    • -
  • Incremental PCA
    • -
  • Randomized PCA
    • -
  • Kernel PCA
    • -
  • LLE
    • -
  • Other techniques
    • +
  • Basic ideas of the Principal Component Analysis (PCA)
    • +
  • Introducing the Covariance and Correlation functions
    • +
  • Classical PCA Theorem
    • +
  • Prof of the PCA Theorem
    • +
  • Getting started with PCA
    • +
  • Principal Component Analysis
    • +
  • PCA and scikit-learn
    • +
  • More on the PCA
    • +
  • Incremental PCA
    • +
  • Randomized PCA
    • +
  • Kernel PCA
    • +
  • LLE
    • +
  • Other techniques
    • @@ -160,7 +174,7 @@ MathJax.Hub.Config({
    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

    -

    Oct 17, 2019

    +

    Oct 19, 2019


    @@ -184,7 +198,7 @@ MathJax.Hub.Config({

  • 9
  • 10
  • ...
    • -
  • 17
    • +
  • 21
  • »
    • diff --git a/doc/pub/DimRed/html/DimRed-reveal.html b/doc/pub/DimRed/html/DimRed-reveal.html index 9fa26b993..82886c900 100644 --- a/doc/pub/DimRed/html/DimRed-reveal.html +++ b/doc/pub/DimRed/html/DimRed-reveal.html @@ -148,7 +148,7 @@ MathJax.Hub.Config({
    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

     
    -

    Oct 17, 2019

    +

    Oct 19, 2019


    @@ -163,16 +163,22 @@ MathJax.Hub.Config({

    -Many Machine Learning problems involve thousands or even millions of features for each training -instance. Not only does this make training extremely slow, it can also make it much harder to find a good -solution, as we will see. This problem is often referred to as the curse of dimensionality. -Fortunately, in real-world problems, it is often possible to reduce the number of features considerably, -turning an intractable problem into a tractable one. +Many Machine Learning problems involve thousands or even millions of +features for each training instance. Not only does this make training +extremely slow, it can also make it much harder to find a good +solution, as we will see. This problem is often referred to as the +curse of dimensionality. Fortunately, in real-world problems, it is +often possible to reduce the number of features considerably, turning +an intractable problem into a tractable one.

    -Here we will discuss some of the most popular dimensionality -reduction techniques: the principal component analysis PCA, Kernel PCA, and Locally Linear Embedding (LLE). -Furthermore, we will start by looking at some simple preprocessing of the data which allow us to rescale the data. +Here we will discuss some of the most popular dimensionality reduction +techniques: the principal component analysis PCA, Kernel PCA, and +Locally Linear Embedding (LLE). Furthermore, we will start by looking +at some simple preprocessing of the data which allow us to rescale the +data. + +

    @@ -183,11 +189,11 @@ Furthermore, we will start by looking at some simple preprocessing of the data w

    Before we proceed however, we will discuss how to preprocess our -data. Till now and in connection with our previous examples we have not met so many cases -where we are too sensitive to the scaling of our data. Normally the -data may need a rescaling and/or may be sensitive to extreme -values. Scaling the data renders our inputs much more suitable for the -algorithms we want to employ. +data. Till now and in connection with our previous examples we have +not met so many cases where we are too sensitive to the scaling of our +data. Normally the data may need a rescaling and/or may be sensitive +to extreme values. Scaling the data renders our inputs much more +suitable for the algorithms we want to employ.

    Scikit-Learn has several functions which allow us to rescale the @@ -482,14 +488,34 @@ X_test_scaled = scaler.transform(X_test) logreg.fit(X_train_scaled, y_train) print("Test set accuracy scaled data: {:.2f}".format(logreg.score(X_test_scaled,y_test)))

    +

    +

    -

    Getting started with PCA

    +

    Basic ideas of the Principal Component Analysis (PCA)

    +
    + + +
    +

    Introducing the Covariance and Correlation functions

    +
    + + +
    +

    Classical PCA Theorem

    +
    + + +
    +

    Prof of the PCA Theorem

    +
    + + +
    +

    Getting started with PCA

    -

    -This material is being finalized, not yet ready.

    @@ -504,7 +530,7 @@ X_pca = pca.transform(X_train_scaled)

    -

    Principal Component Analysis

    +

    Principal Component Analysis

    @@ -541,7 +567,7 @@ X2D = X_centered.dot(W2)

    -

    PCA and scikit-learn

    +

    PCA and scikit-learn

    Scikit-Learn’s PCA class implements PCA using SVD decomposition just like we did before. The @@ -572,7 +598,9 @@ More material to come here.

    -

    More on the PCA

    +

    More on the PCA

    + +

    Instead of arbitrarily choosing the number of dimensions to reduce down to, it is generally preferable to choose the number of dimensions that add up to a sufficiently large portion of the variance (e.g., 95%). Unless, of course, you are reducing dimensionality for data visualization — in that case you will @@ -601,7 +629,9 @@ X_reduced = pca.fit_transform(X)

    -

    Incremental PCA

    +

    Incremental PCA

    + +

    One problem with the preceding implementation of PCA is that it requires the whole training set to fit in memory in order for the SVD algorithm to run. Fortunately, Incremental PCA (IPCA) algorithms have been developed: you can split the training set into mini-batches and feed an IPCA algorithm one minibatch @@ -611,7 +641,7 @@ instances arrive).

    -

    Randomized PCA

    +

    Randomized PCA

    Scikit-Learn offers yet another option to perform PCA, called Randomized PCA. This is a stochastic @@ -625,7 +655,7 @@ previous algorithms when \( d \) is much smaller than \( n \).

    -

    Kernel PCA

    +

    Kernel PCA

    @@ -651,7 +681,7 @@ X_reduced = rbf_pca.fit_transform(X)

    -

    LLE

    +

    LLE

    Locally Linear Embedding (LLE) is another very powerful nonlinear dimensionality reduction @@ -663,7 +693,7 @@ these local relationships are best preserved (more details shortly).

    -

    Other techniques

    +

    Other techniques

    There are many other dimensionality reduction techniques, several of which are available in Scikit-Learn. diff --git a/doc/pub/DimRed/html/DimRed-solarized.html b/doc/pub/DimRed/html/DimRed-solarized.html index 6b682f7ea..f777e2e44 100644 --- a/doc/pub/DimRed/html/DimRed-solarized.html +++ b/doc/pub/DimRed/html/DimRed-solarized.html @@ -84,15 +84,25 @@ div { text-align: justify; text-justify: inter-word; } 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -134,7 +144,7 @@ MathJax.Hub.Config({

    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

    -

    Oct 17, 2019

    +

    Oct 19, 2019












    @@ -145,16 +155,22 @@ MathJax.Hub.Config({

    -Many Machine Learning problems involve thousands or even millions of features for each training -instance. Not only does this make training extremely slow, it can also make it much harder to find a good -solution, as we will see. This problem is often referred to as the curse of dimensionality. -Fortunately, in real-world problems, it is often possible to reduce the number of features considerably, -turning an intractable problem into a tractable one. +Many Machine Learning problems involve thousands or even millions of +features for each training instance. Not only does this make training +extremely slow, it can also make it much harder to find a good +solution, as we will see. This problem is often referred to as the +curse of dimensionality. Fortunately, in real-world problems, it is +often possible to reduce the number of features considerably, turning +an intractable problem into a tractable one.

    -Here we will discuss some of the most popular dimensionality -reduction techniques: the principal component analysis PCA, Kernel PCA, and Locally Linear Embedding (LLE). -Furthermore, we will start by looking at some simple preprocessing of the data which allow us to rescale the data. +Here we will discuss some of the most popular dimensionality reduction +techniques: the principal component analysis PCA, Kernel PCA, and +Locally Linear Embedding (LLE). Furthermore, we will start by looking +at some simple preprocessing of the data which allow us to rescale the +data. + +

    @@ -168,11 +184,11 @@ Furthermore, we will start by looking at some simple preprocessing of the data w

    Before we proceed however, we will discuss how to preprocess our -data. Till now and in connection with our previous examples we have not met so many cases -where we are too sensitive to the scaling of our data. Normally the -data may need a rescaling and/or may be sensitive to extreme -values. Scaling the data renders our inputs much more suitable for the -algorithms we want to employ. +data. Till now and in connection with our previous examples we have +not met so many cases where we are too sensitive to the scaling of our +data. Normally the data may need a rescaling and/or may be sensitive +to extreme values. Scaling the data renders our inputs much more +suitable for the algorithms we want to employ.

    Scikit-Learn has several functions which allow us to rescale the @@ -466,12 +482,33 @@ logreg.fit(X_train_scaled, y_train) print("Test set accuracy scaled data: {:.2f}".format(logreg.score(X_test_scaled,y_test)))

    -









    - -

    Getting started with PCA

    +

    -This material is being finalized, not yet ready. +









    + +

    Basic ideas of the Principal Component Analysis (PCA)

    + +

    +









    + +

    Introducing the Covariance and Correlation functions

    + +

    +









    + +

    Classical PCA Theorem

    + +

    +









    + +

    Prof of the PCA Theorem

    + +

    +









    + +

    Getting started with PCA

    +

    @@ -485,7 +522,7 @@ X_pca = pca.transform(X_train_scaled)











    -

    Principal Component Analysis

    +

    Principal Component Analysis

    @@ -521,7 +558,7 @@ X2D = X_centered.dot(W2)

    -

    PCA and scikit-learn

    +

    PCA and scikit-learn

    Scikit-Learn’s PCA class implements PCA using SVD decomposition just like we did before. The @@ -552,7 +589,9 @@ More material to come here.











    -

    More on the PCA

    +

    More on the PCA

    + +

    Instead of arbitrarily choosing the number of dimensions to reduce down to, it is generally preferable to choose the number of dimensions that add up to a sufficiently large portion of the variance (e.g., 95%). Unless, of course, you are reducing dimensionality for data visualization — in that case you will @@ -580,7 +619,9 @@ X_reduced = pca.fit_transform(X)











    -

    Incremental PCA

    +

    Incremental PCA

    + +

    One problem with the preceding implementation of PCA is that it requires the whole training set to fit in memory in order for the SVD algorithm to run. Fortunately, Incremental PCA (IPCA) algorithms have been developed: you can split the training set into mini-batches and feed an IPCA algorithm one minibatch @@ -590,7 +631,7 @@ instances arrive).











    -

    Randomized PCA

    +

    Randomized PCA

    Scikit-Learn offers yet another option to perform PCA, called Randomized PCA. This is a stochastic @@ -605,7 +646,7 @@ previous algorithms when \( d \) is much smaller than \( n \).











    -

    Kernel PCA

    +

    Kernel PCA

    @@ -634,7 +675,7 @@ X_reduced = rbf_pca.fit_transform(X)











    -

    LLE

    +

    LLE

    Locally Linear Embedding (LLE) is another very powerful nonlinear dimensionality reduction @@ -646,7 +687,7 @@ these local relationships are best preserved (more details shortly).











    -

    Other techniques

    +

    Other techniques

    There are many other dimensionality reduction techniques, several of which are available in Scikit-Learn. diff --git a/doc/pub/DimRed/html/DimRed.html b/doc/pub/DimRed/html/DimRed.html index 5b3216924..16c9a74ea 100644 --- a/doc/pub/DimRed/html/DimRed.html +++ b/doc/pub/DimRed/html/DimRed.html @@ -89,15 +89,25 @@ div { text-align: justify; text-justify: inter-word; } 2, None, '___sec6'), - ('Getting started with PCA', 2, None, '___sec7'), - ('Principal Component Analysis', 2, None, '___sec8'), - ('PCA and scikit-learn', 2, None, '___sec9'), - ('More on the PCA', 2, None, '___sec10'), - ('Incremental PCA', 2, None, '___sec11'), - ('Randomized PCA', 2, None, '___sec12'), - ('Kernel PCA', 2, None, '___sec13'), - ('LLE', 2, None, '___sec14'), - ('Other techniques', 2, None, '___sec15')]} + ('Basic ideas of the Principal Component Analysis (PCA)', + 2, + None, + '___sec7'), + ('Introducing the Covariance and Correlation functions', + 2, + None, + '___sec8'), + ('Classical PCA Theorem', 2, None, '___sec9'), + ('Prof of the PCA Theorem', 2, None, '___sec10'), + ('Getting started with PCA', 2, None, '___sec11'), + ('Principal Component Analysis', 2, None, '___sec12'), + ('PCA and scikit-learn', 2, None, '___sec13'), + ('More on the PCA', 2, None, '___sec14'), + ('Incremental PCA', 2, None, '___sec15'), + ('Randomized PCA', 2, None, '___sec16'), + ('Kernel PCA', 2, None, '___sec17'), + ('LLE', 2, None, '___sec18'), + ('Other techniques', 2, None, '___sec19')]} end of tocinfo --> @@ -139,7 +149,7 @@ MathJax.Hub.Config({

    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

    -

    Oct 17, 2019

    +

    Oct 19, 2019












    @@ -150,16 +160,22 @@ MathJax.Hub.Config({

    -Many Machine Learning problems involve thousands or even millions of features for each training -instance. Not only does this make training extremely slow, it can also make it much harder to find a good -solution, as we will see. This problem is often referred to as the curse of dimensionality. -Fortunately, in real-world problems, it is often possible to reduce the number of features considerably, -turning an intractable problem into a tractable one. +Many Machine Learning problems involve thousands or even millions of +features for each training instance. Not only does this make training +extremely slow, it can also make it much harder to find a good +solution, as we will see. This problem is often referred to as the +curse of dimensionality. Fortunately, in real-world problems, it is +often possible to reduce the number of features considerably, turning +an intractable problem into a tractable one.

    -Here we will discuss some of the most popular dimensionality -reduction techniques: the principal component analysis PCA, Kernel PCA, and Locally Linear Embedding (LLE). -Furthermore, we will start by looking at some simple preprocessing of the data which allow us to rescale the data. +Here we will discuss some of the most popular dimensionality reduction +techniques: the principal component analysis PCA, Kernel PCA, and +Locally Linear Embedding (LLE). Furthermore, we will start by looking +at some simple preprocessing of the data which allow us to rescale the +data. + +

    @@ -173,11 +189,11 @@ Furthermore, we will start by looking at some simple preprocessing of the data w

    Before we proceed however, we will discuss how to preprocess our -data. Till now and in connection with our previous examples we have not met so many cases -where we are too sensitive to the scaling of our data. Normally the -data may need a rescaling and/or may be sensitive to extreme -values. Scaling the data renders our inputs much more suitable for the -algorithms we want to employ. +data. Till now and in connection with our previous examples we have +not met so many cases where we are too sensitive to the scaling of our +data. Normally the data may need a rescaling and/or may be sensitive +to extreme values. Scaling the data renders our inputs much more +suitable for the algorithms we want to employ.

    Scikit-Learn has several functions which allow us to rescale the @@ -471,12 +487,33 @@ logreg.fit(X_train_scaled, y_train) print("Test set accuracy scaled data: {:.2f}".format(logreg.score(X_test_scaled,y_test)))

    -









    - -

    Getting started with PCA

    +

    -This material is being finalized, not yet ready. +









    + +

    Basic ideas of the Principal Component Analysis (PCA)

    + +

    +









    + +

    Introducing the Covariance and Correlation functions

    + +

    +









    + +

    Classical PCA Theorem

    + +

    +









    + +

    Prof of the PCA Theorem

    + +

    +









    + +

    Getting started with PCA

    +

    @@ -490,7 +527,7 @@ X_pca = pca.









    -

    Principal Component Analysis

    +

    Principal Component Analysis

    @@ -526,7 +563,7 @@ X2D = X_centered -

    PCA and scikit-learn

    +

    PCA and scikit-learn

    Scikit-Learn’s PCA class implements PCA using SVD decomposition just like we did before. The @@ -557,7 +594,9 @@ More material to come here.











    -

    More on the PCA

    +

    More on the PCA

    + +

    Instead of arbitrarily choosing the number of dimensions to reduce down to, it is generally preferable to choose the number of dimensions that add up to a sufficiently large portion of the variance (e.g., 95%). Unless, of course, you are reducing dimensionality for data visualization — in that case you will @@ -585,7 +624,9 @@ X_reduced = pca











    -

    Incremental PCA

    +

    Incremental PCA

    + +

    One problem with the preceding implementation of PCA is that it requires the whole training set to fit in memory in order for the SVD algorithm to run. Fortunately, Incremental PCA (IPCA) algorithms have been developed: you can split the training set into mini-batches and feed an IPCA algorithm one minibatch @@ -595,7 +636,7 @@ instances arrive).











    -

    Randomized PCA

    +

    Randomized PCA

    Scikit-Learn offers yet another option to perform PCA, called Randomized PCA. This is a stochastic @@ -610,7 +651,7 @@ previous algorithms when \( d \) is much smaller than \( n \).











    -

    Kernel PCA

    +

    Kernel PCA

    @@ -639,7 +680,7 @@ X_reduced = rbf_pcaLLE +

    LLE

    Locally Linear Embedding (LLE) is another very powerful nonlinear dimensionality reduction @@ -651,7 +692,7 @@ these local relationships are best preserved (more details shortly).











    -

    Other techniques

    +

    Other techniques

    There are many other dimensionality reduction techniques, several of which are available in Scikit-Learn. diff --git a/doc/pub/DimRed/ipynb/DimRed.ipynb b/doc/pub/DimRed/ipynb/DimRed.ipynb index 3a506726d..cc73fb1ff 100644 --- a/doc/pub/DimRed/ipynb/DimRed.ipynb +++ b/doc/pub/DimRed/ipynb/DimRed.ipynb @@ -10,7 +10,7 @@ " \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 17, 2019**\n", + "Date: **Oct 19, 2019**\n", "\n", "Copyright 1999-2019, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license\n", "\n", @@ -19,15 +19,19 @@ "\n", "## Reducing the number of degrees of freedom, overarching view\n", "\n", - "Many Machine Learning problems involve thousands or even millions of features for each training\n", - "instance. Not only does this make training extremely slow, it can also make it much harder to find a good\n", - "solution, as we will see. This problem is often referred to as the curse of dimensionality.\n", - "Fortunately, in real-world problems, it is often possible to reduce the number of features considerably,\n", - "turning an intractable problem into a tractable one.\n", + "Many Machine Learning problems involve thousands or even millions of\n", + "features for each training instance. Not only does this make training\n", + "extremely slow, it can also make it much harder to find a good\n", + "solution, as we will see. This problem is often referred to as the\n", + "curse of dimensionality. Fortunately, in real-world problems, it is\n", + "often possible to reduce the number of features considerably, turning\n", + "an intractable problem into a tractable one.\n", "\n", - "Here we will discuss some of the most popular dimensionality\n", - "reduction techniques: the principal component analysis PCA, Kernel PCA, and Locally Linear Embedding (LLE).\n", - "Furthermore, we will start by looking at some simple preprocessing of the data which allow us to rescale the data.\n", + "Here we will discuss some of the most popular dimensionality reduction\n", + "techniques: the principal component analysis PCA, Kernel PCA, and\n", + "Locally Linear Embedding (LLE). Furthermore, we will start by looking\n", + "at some simple preprocessing of the data which allow us to rescale the\n", + "data.\n", "\n", "\n", "\n", @@ -35,11 +39,11 @@ "## Preprocessing our data\n", "\n", "Before we proceed however, we will discuss how to preprocess our\n", - "data. Till now and in connection with our previous examples we have not met so many cases\n", - "where we are too sensitive to the scaling of our data. Normally the\n", - "data may need a rescaling and/or may be sensitive to extreme\n", - "values. Scaling the data renders our inputs much more suitable for the\n", - "algorithms we want to employ.\n", + "data. Till now and in connection with our previous examples we have\n", + "not met so many cases where we are too sensitive to the scaling of our\n", + "data. Normally the data may need a rescaling and/or may be sensitive\n", + "to extreme values. Scaling the data renders our inputs much more\n", + "suitable for the algorithms we want to employ.\n", "\n", "**Scikit-Learn** has several functions which allow us to rescale the\n", "data, normally resulting in much better results in terms of various\n", @@ -79,41 +83,10 @@ { "cell_type": "code", "execution_count": 1, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "MSE before scaling: 0.01\n", - "R2 score before scaling 0.93\n", - "Feature min values before scaling:\n", - " [1.00000000e+00 1.56868531e-04 4.25016115e-04 2.46077359e-08\n", - " 6.66716535e-08 1.80638698e-07 3.86017938e-12 1.04586843e-11\n", - " 2.83365271e-11 7.67743576e-11 6.05540668e-16 1.64063844e-15\n", - " 4.44510938e-15 1.20434807e-14 3.26303392e-14 9.49902749e-20\n", - " 2.57364542e-19 6.97297777e-19 1.88924312e-18 5.11867337e-18\n", - " 1.38684200e-17]\n", - "Feature max values before scaling:\n", - " [1. 0.99980229 0.9996158 0.99960462 0.99941816 0.99923174\n", - " 0.99940699 0.99922057 0.99903418 0.99884783 0.99920939 0.99902301\n", - " 0.99883666 0.99865035 0.99846407 0.99901184 0.99882549 0.99863918\n", - " 0.99845291 0.99826666 0.99808046]\n", - "Feature min values after scaling:\n", - " [ 0. -1.78024859 -1.70920112 -1.1515339 -1.1300959 -1.11033121\n", - " -0.90340545 -0.8959348 -0.88900131 -0.88255214 -0.76338933 -0.76113708\n", - " -0.75906691 -0.7571601 -0.75539963 -0.67113296 -0.67101229 -0.6709573\n", - " -0.67096075 -0.67101596 -0.67111677]\n", - "Feature max values after scaling:\n", - " [0. 1.74136774 1.71581105 2.27322963 2.24138196 2.20981521\n", - " 2.70775213 2.67747156 2.64735409 2.61738276 3.08600037 3.05801994\n", - " 3.03002682 3.00201047 2.97396141 3.42593492 3.39990873 3.37378657\n", - " 3.34756305 3.32123338 3.29479332]\n", - "MSE after scaling: 0.00\n", - "R2 score for scaled data: 0.97\n" - ] - } - ], + "metadata": { + "collapsed": false + }, + "outputs": [], "source": [ "%matplotlib inline\n", "\n", @@ -227,48 +200,10 @@ { "cell_type": "code", "execution_count": 2, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "(426, 30)\n", - "(143, 30)\n", - "Test set accuracy: 0.63\n", - "Feature min values before scaling:\n", - " [6.981e+00 9.710e+00 4.379e+01 1.435e+02 5.263e-02 1.938e-02 0.000e+00\n", - " 0.000e+00 1.060e-01 4.996e-02 1.115e-01 3.628e-01 7.570e-01 7.228e+00\n", - " 1.713e-03 2.252e-03 0.000e+00 0.000e+00 7.882e-03 8.948e-04 7.930e+00\n", - " 1.202e+01 5.041e+01 1.852e+02 7.117e-02 2.729e-02 0.000e+00 0.000e+00\n", - " 1.565e-01 5.504e-02]\n", - "Feature max values before scaling:\n", - " [2.811e+01 3.381e+01 1.885e+02 2.501e+03 1.447e-01 3.114e-01 4.268e-01\n", - " 2.012e-01 3.040e-01 9.744e-02 2.873e+00 4.885e+00 2.198e+01 5.422e+02\n", - " 2.333e-02 1.064e-01 3.960e-01 5.279e-02 6.146e-02 2.984e-02 3.604e+01\n", - " 4.954e+01 2.512e+02 4.254e+03 2.226e-01 1.058e+00 1.252e+00 2.903e-01\n", - " 6.638e-01 2.075e-01]\n", - "Feature min values before scaling:\n", - " [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n", - " 0. 0. 0. 0. 0. 0.]\n", - "Feature max values before scaling:\n", - " [1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n", - " 1. 1. 1. 1. 1. 1.]\n", - "Test set accuracy scaled data with Min-Max scaling: 0.97\n", - "Test set accuracy scaled data with Standar Scaler: 0.96\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/usr/local/lib/python3.7/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n", - " \"avoid this warning.\", FutureWarning)\n", - "/usr/local/lib/python3.7/site-packages/sklearn/svm/base.py:196: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n", - " \"avoid this warning.\", FutureWarning)\n" - ] - } - ], + "metadata": { + "collapsed": false + }, + "outputs": [], "source": [ "import matplotlib.pyplot as plt\n", "import numpy as np\n", @@ -320,7 +255,9 @@ { "cell_type": "code", "execution_count": 3, - "metadata": {}, + "metadata": { + "collapsed": false + }, "outputs": [], "source": [ "import matplotlib.pyplot as plt\n", @@ -358,58 +295,11 @@ }, { "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [ - { - "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" - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "[ 1.33026664e+01 5.69238112e+00 2.83259341e+00 1.99126621e+00\n", - " 1.69095941e+00 1.19614372e+00 7.13037918e-01 5.77804918e-01\n", - " 5.14549644e-01 4.30579019e-01 3.58977964e-01 -1.91269450e-01\n", - " 2.86195387e-01 -1.59310961e-01 -1.38483045e-01 2.44786221e-01\n", - " -8.79318674e-02 1.91811329e-01 1.70896266e-01 -7.17683028e-02\n", - " 1.39860289e-01 1.18721828e-01 -4.83194721e-02 8.92983892e-02\n", - " 6.81402316e-02 6.01968016e-02 3.31819744e-02 1.56432255e-02\n", - " -1.61028757e-02 -6.50565973e-03]\n", - "(426, 30)\n", - "(143, 30)\n", - "Test set accuracy from Logistic Regression: 0.96\n", - "Test set accuracy scaled data: 0.96\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/usr/local/lib/python3.7/site-packages/sklearn/linear_model/logistic.py:432: FutureWarning: Default solver will be changed to 'lbfgs' in 0.22. Specify a solver to silence this warning.\n", - " FutureWarning)\n", - "/usr/local/lib/python3.7/site-packages/sklearn/linear_model/logistic.py:432: FutureWarning: Default solver will be changed to 'lbfgs' in 0.22. Specify a solver to silence this warning.\n", - " FutureWarning)\n" - ] - } - ], + "execution_count": 4, + "metadata": { + "collapsed": false + }, + "outputs": [], "source": [ "import matplotlib.pyplot as plt\n", "import numpy as np\n", @@ -472,15 +362,33 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "## Getting started with PCA\n", + "\n", "\n", - "This material is being finalized, not yet ready." + "\n", + "## Basic ideas of the Principal Component Analysis (PCA)\n", + "\n", + "\n", + "## Introducing the Covariance and Correlation functions\n", + "\n", + "## Classical PCA Theorem\n", + "\n", + "\n", + "## Prof of the PCA Theorem\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "## Getting started with PCA" ] }, { "cell_type": "code", "execution_count": 5, - "metadata": {}, + "metadata": { + "collapsed": false + }, "outputs": [], "source": [ "# Now add PCA\n", @@ -506,7 +414,9 @@ { "cell_type": "code", "execution_count": 6, - "metadata": {}, + "metadata": { + "collapsed": false + }, "outputs": [], "source": [ "X_centered = X - X.mean(axis=0)\n", @@ -531,7 +441,9 @@ { "cell_type": "code", "execution_count": 7, - "metadata": {}, + "metadata": { + "collapsed": false + }, "outputs": [], "source": [ "W2 = V.T[:, :2]\n", @@ -553,7 +465,9 @@ { "cell_type": "code", "execution_count": 8, - "metadata": {}, + "metadata": { + "collapsed": false + }, "outputs": [], "source": [ "from sklearn.decomposition import PCA\n", @@ -573,7 +487,9 @@ { "cell_type": "code", "execution_count": 9, - "metadata": {}, + "metadata": { + "collapsed": false + }, "outputs": [], "source": [ "pca.components_.T[:, 0])." @@ -589,6 +505,7 @@ "More material to come here.\n", "\n", "## More on the PCA\n", + "\n", "Instead of arbitrarily choosing the number of dimensions to reduce down to, it is generally preferable to\n", "choose the number of dimensions that add up to a sufficiently large portion of the variance (e.g., 95%).\n", "Unless, of course, you are reducing dimensionality for data visualization — in that case you will\n", @@ -600,7 +517,9 @@ { "cell_type": "code", "execution_count": 10, - "metadata": {}, + "metadata": { + "collapsed": false + }, "outputs": [], "source": [ "pca = PCA()\n", @@ -621,7 +540,9 @@ { "cell_type": "code", "execution_count": 11, - "metadata": {}, + "metadata": { + "collapsed": false + }, "outputs": [], "source": [ "pca = PCA(n_components=0.95)\n", @@ -633,6 +554,7 @@ "metadata": {}, "source": [ "## Incremental PCA\n", + "\n", "One problem with the preceding implementation of PCA is that it requires the whole training set to fit in\n", "memory in order for the SVD algorithm to run. Fortunately, Incremental PCA (IPCA) algorithms have\n", "been developed: you can split the training set into mini-batches and feed an IPCA algorithm one minibatch\n", @@ -665,7 +587,9 @@ { "cell_type": "code", "execution_count": 12, - "metadata": {}, + "metadata": { + "collapsed": false + }, "outputs": [], "source": [ "from sklearn.decomposition import KernelPCA\n", @@ -703,25 +627,7 @@ ] } ], - "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": "3.7.4" - } - }, + "metadata": {}, "nbformat": 4, "nbformat_minor": 2 } diff --git a/doc/pub/DimRed/ipynb/ipynb-DimRed-src.tar.gz b/doc/pub/DimRed/ipynb/ipynb-DimRed-src.tar.gz index 2d3fa6507..cc8d3012b 100644 Binary files a/doc/pub/DimRed/ipynb/ipynb-DimRed-src.tar.gz and b/doc/pub/DimRed/ipynb/ipynb-DimRed-src.tar.gz differ diff --git a/doc/pub/DimRed/pdf/DimRed-minted.pdf b/doc/pub/DimRed/pdf/DimRed-minted.pdf index 6bbfbdce6..85a5cc870 100644 Binary files a/doc/pub/DimRed/pdf/DimRed-minted.pdf and b/doc/pub/DimRed/pdf/DimRed-minted.pdf differ diff --git a/doc/src/DimRed/DimRed.do.txt b/doc/src/DimRed/DimRed.do.txt index 536a45c0f..b7569082c 100644 --- a/doc/src/DimRed/DimRed.do.txt +++ b/doc/src/DimRed/DimRed.do.txt @@ -7,15 +7,21 @@ DATE: today ===== Reducing the number of degrees of freedom, overarching view ===== !bblock -Many Machine Learning problems involve thousands or even millions of features for each training -instance. Not only does this make training extremely slow, it can also make it much harder to find a good -solution, as we will see. This problem is often referred to as the curse of dimensionality. -Fortunately, in real-world problems, it is often possible to reduce the number of features considerably, -turning an intractable problem into a tractable one. +Many Machine Learning problems involve thousands or even millions of +features for each training instance. Not only does this make training +extremely slow, it can also make it much harder to find a good +solution, as we will see. This problem is often referred to as the +curse of dimensionality. Fortunately, in real-world problems, it is +often possible to reduce the number of features considerably, turning +an intractable problem into a tractable one. + +Here we will discuss some of the most popular dimensionality reduction +techniques: the principal component analysis PCA, Kernel PCA, and +Locally Linear Embedding (LLE). Furthermore, we will start by looking +at some simple preprocessing of the data which allow us to rescale the +data. + -Here we will discuss some of the most popular dimensionality -reduction techniques: the principal component analysis PCA, Kernel PCA, and Locally Linear Embedding (LLE). -Furthermore, we will start by looking at some simple preprocessing of the data which allow us to rescale the data. !eblock @@ -24,11 +30,11 @@ Furthermore, we will start by looking at some simple preprocessing of the data w !bblock Before we proceed however, we will discuss how to preprocess our -data. Till now and in connection with our previous examples we have not met so many cases -where we are too sensitive to the scaling of our data. Normally the -data may need a rescaling and/or may be sensitive to extreme -values. Scaling the data renders our inputs much more suitable for the -algorithms we want to employ. +data. Till now and in connection with our previous examples we have +not met so many cases where we are too sensitive to the scaling of our +data. Normally the data may need a rescaling and/or may be sensitive +to extreme values. Scaling the data renders our inputs much more +suitable for the algorithms we want to employ. _Scikit-Learn_ has several functions which allow us to rescale the data, normally resulting in much better results in terms of various @@ -310,18 +316,36 @@ X_test_scaled = scaler.transform(X_test) logreg.fit(X_train_scaled, y_train) print("Test set accuracy scaled data: {:.2f}".format(logreg.score(X_test_scaled,y_test))) - - - - !ec +#todo: add more text in order to explain what is done here, discuss the correlation matrix + + +!split +===== Basic ideas of the Principal Component Analysis (PCA) ===== + + +!split +===== Introducing the Covariance and Correlation functions ===== + +!split +===== Classical PCA Theorem ===== + + +!split +===== Prof of the PCA Theorem ===== + + + + + + !split ===== Getting started with PCA ===== -This material is being finalized, not yet ready. + !bc pycod # Now add PCA from sklearn.decomposition import PCA @@ -384,6 +408,7 @@ More material to come here. !split ===== More on the PCA ===== + Instead of arbitrarily choosing the number of dimensions to reduce down to, it is generally preferable to choose the number of dimensions that add up to a sufficiently large portion of the variance (e.g., 95%). Unless, of course, you are reducing dimensionality for data visualization — in that case you will @@ -406,6 +431,7 @@ X_reduced = pca.fit_transform(X) !split ===== Incremental PCA ===== + One problem with the preceding implementation of PCA is that it requires the whole training set to fit in memory in order for the SVD algorithm to run. Fortunately, Incremental PCA (IPCA) algorithms have been developed: you can split the training set into mini-batches and feed an IPCA algorithm one minibatch