Files
FYS-STK4155/doc/pub/week39/html/._week39-bs060.html
T
2020-09-24 06:54:17 +02:00

396 lines
24 KiB
HTML

<!--
Automatically generated HTML file from DocOnce source
(https://github.com/hplgit/doconce/)
-->
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<meta name="generator" content="DocOnce: https://github.com/hplgit/doconce/" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<meta name="description" content="Week 39: Optimization and Gradient Methods">
<title>Week 39: Optimization and Gradient Methods</title>
<!-- Bootstrap style: bootstrap -->
<link href="https://netdna.bootstrapcdn.com/bootstrap/3.1.1/css/bootstrap.min.css" rel="stylesheet">
<!-- not necessary
<link href="https://netdna.bootstrapcdn.com/font-awesome/4.0.3/css/font-awesome.css" rel="stylesheet">
-->
<style type="text/css">
/* Add scrollbar to dropdown menus in bootstrap navigation bar */
.dropdown-menu {
height: auto;
max-height: 400px;
overflow-x: hidden;
}
/* Adds an invisible element before each target to offset for the navigation
bar */
.anchor::before {
content:"";
display:block;
height:50px; /* fixed header height for style bootstrap */
margin:-50px 0 0; /* negative fixed header height */
}
</style>
</head>
<!-- tocinfo
{'highest level': 2,
'sections': [('Plan for week 39', 2, None, '___sec0'),
('Thursday September 24', 2, None, '___sec1'),
('Optimization, the central part of any Machine Learning '
'algortithm',
2,
None,
'___sec2'),
('Revisiting our Logistic Regression case', 2, None, '___sec3'),
('The equations to solve', 2, None, '___sec4'),
("Solving using Newton-Raphson's method", 2, None, '___sec5'),
("Brief reminder on Newton-Raphson's method", 2, None, '___sec6'),
('The equations', 2, None, '___sec7'),
('Simple geometric interpretation', 2, None, '___sec8'),
('Extending to more than one variable', 2, None, '___sec9'),
('Steepest descent', 2, None, '___sec10'),
('More on Steepest descent', 2, None, '___sec11'),
('The ideal', 2, None, '___sec12'),
('The sensitiveness of the gradient descent',
2,
None,
'___sec13'),
('Convex functions', 2, None, '___sec14'),
('Convex function', 2, None, '___sec15'),
('Conditions on convex functions', 2, None, '___sec16'),
('More on convex functions', 2, None, '___sec17'),
('Some simple problems', 2, None, '___sec18'),
('Standard steepest descent', 2, None, '___sec19'),
('Gradient method', 2, None, '___sec20'),
('Steepest descent method', 2, None, '___sec21'),
('Steepest descent method', 2, None, '___sec22'),
('Final expressions', 2, None, '___sec23'),
('Steepest descent example', 2, None, '___sec24'),
('Conjugate gradient method', 2, None, '___sec25'),
('Conjugate gradient method', 2, None, '___sec26'),
('Conjugate gradient method', 2, None, '___sec27'),
('Conjugate gradient method', 2, None, '___sec28'),
('Conjugate gradient method and iterations', 2, None, '___sec29'),
('Conjugate gradient method', 2, None, '___sec30'),
('Conjugate gradient method', 2, None, '___sec31'),
('Conjugate gradient method', 2, None, '___sec32'),
('Revisiting our first homework', 2, None, '___sec33'),
('Gradient descent example', 2, None, '___sec34'),
('The derivative of the cost/loss function', 2, None, '___sec35'),
('The Hessian matrix', 2, None, '___sec36'),
('Simple program', 2, None, '___sec37'),
('Gradient Descent Example', 2, None, '___sec38'),
('And a corresponding example using _scikit-learn_',
2,
None,
'___sec39'),
('Gradient descent and Ridge', 2, None, '___sec40'),
('Program example for gradient descent with Ridge Regression',
2,
None,
'___sec41'),
('Using gradient descent methods, limitations',
2,
None,
'___sec42'),
('Friday September 25', 2, None, '___sec43'),
('Stochastic Gradient Descent', 2, None, '___sec44'),
('Computation of gradients', 2, None, '___sec45'),
('SGD example', 2, None, '___sec46'),
('The gradient step', 2, None, '___sec47'),
('Simple example code', 2, None, '___sec48'),
('When do we stop?', 2, None, '___sec49'),
('Slightly different approach', 2, None, '___sec50'),
('Program for stochastic gradient', 2, None, '___sec51'),
('Momentum based GD', 2, None, '___sec52'),
('More on momentum based approaches', 2, None, '___sec53'),
('Momentum parameter', 2, None, '___sec54'),
('Second moment of the gradient', 2, None, '___sec55'),
('RMS prop', 2, None, '___sec56'),
('ADAM optimizer', 2, None, '___sec57'),
('Practical tips', 2, None, '___sec58'),
('Automatic differentiation', 2, None, '___sec59'),
('Using autograd', 2, None, '___sec60'),
('Autograd with more complicated functions', 2, None, '___sec61'),
('More complicated functions using the elements of their '
'arguments directly',
2,
None,
'___sec62'),
('Functions using mathematical functions from Numpy',
2,
None,
'___sec63'),
('More autograd', 2, None, '___sec64'),
('And with loops', 2, None, '___sec65'),
('Using recursion', 2, None, '___sec66'),
('Unsupported functions', 2, None, '___sec67'),
('The syntax a.dot(b) when finding the dot product',
2,
None,
'___sec68'),
('Recommended to avoid', 2, None, '___sec69')]}
end of tocinfo -->
<body>
<script type="text/x-mathjax-config">
MathJax.Hub.Config({
TeX: {
equationNumbers: { autoNumber: "none" },
extensions: ["AMSmath.js", "AMSsymbols.js", "autobold.js", "color.js"]
}
});
</script>
<script type="text/javascript" async
src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.1/MathJax.js?config=TeX-AMS-MML_HTMLorMML">
</script>
<!-- Bootstrap navigation bar -->
<div class="navbar navbar-default navbar-fixed-top">
<div class="navbar-header">
<button type="button" class="navbar-toggle" data-toggle="collapse" data-target=".navbar-responsive-collapse">
<span class="icon-bar"></span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
</button>
<a class="navbar-brand" href="week39-bs.html">Week 39: Optimization and Gradient Methods</a>
</div>
<div class="navbar-collapse collapse navbar-responsive-collapse">
<ul class="nav navbar-nav navbar-right">
<li class="dropdown">
<a href="#" class="dropdown-toggle" data-toggle="dropdown">Contents <b class="caret"></b></a>
<ul class="dropdown-menu">
<!-- navigation toc: --> <li><a href="._week39-bs001.html#___sec0" style="font-size: 80%;">Plan for week 39</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs002.html#___sec1" style="font-size: 80%;">Thursday September 24</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs003.html#___sec2" style="font-size: 80%;">Optimization, the central part of any Machine Learning algortithm</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs004.html#___sec3" style="font-size: 80%;">Revisiting our Logistic Regression case</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs005.html#___sec4" style="font-size: 80%;">The equations to solve</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs006.html#___sec5" style="font-size: 80%;">Solving using Newton-Raphson's method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs007.html#___sec6" style="font-size: 80%;">Brief reminder on Newton-Raphson's method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs008.html#___sec7" style="font-size: 80%;">The equations</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs009.html#___sec8" style="font-size: 80%;">Simple geometric interpretation</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs010.html#___sec9" style="font-size: 80%;">Extending to more than one variable</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs011.html#___sec10" style="font-size: 80%;">Steepest descent</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs012.html#___sec11" style="font-size: 80%;">More on Steepest descent</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs013.html#___sec12" style="font-size: 80%;">The ideal</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs014.html#___sec13" style="font-size: 80%;">The sensitiveness of the gradient descent</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs015.html#___sec14" style="font-size: 80%;">Convex functions</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs016.html#___sec15" style="font-size: 80%;">Convex function</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs017.html#___sec16" style="font-size: 80%;">Conditions on convex functions</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs018.html#___sec17" style="font-size: 80%;">More on convex functions</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs019.html#___sec18" style="font-size: 80%;">Some simple problems</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs020.html#___sec19" style="font-size: 80%;">Standard steepest descent</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs021.html#___sec20" style="font-size: 80%;">Gradient method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs022.html#___sec21" style="font-size: 80%;">Steepest descent method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs023.html#___sec22" style="font-size: 80%;">Steepest descent method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs024.html#___sec23" style="font-size: 80%;">Final expressions</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs025.html#___sec24" style="font-size: 80%;">Steepest descent example</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs026.html#___sec25" style="font-size: 80%;">Conjugate gradient method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs027.html#___sec26" style="font-size: 80%;">Conjugate gradient method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs028.html#___sec27" style="font-size: 80%;">Conjugate gradient method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs029.html#___sec28" style="font-size: 80%;">Conjugate gradient method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs030.html#___sec29" style="font-size: 80%;">Conjugate gradient method and iterations</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs031.html#___sec30" style="font-size: 80%;">Conjugate gradient method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs032.html#___sec31" style="font-size: 80%;">Conjugate gradient method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs033.html#___sec32" style="font-size: 80%;">Conjugate gradient method</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs034.html#___sec33" style="font-size: 80%;">Revisiting our first homework</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs035.html#___sec34" style="font-size: 80%;">Gradient descent example</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs036.html#___sec35" style="font-size: 80%;">The derivative of the cost/loss function</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs037.html#___sec36" style="font-size: 80%;">The Hessian matrix</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs038.html#___sec37" style="font-size: 80%;">Simple program</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs039.html#___sec38" style="font-size: 80%;">Gradient Descent Example</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs040.html#___sec39" style="font-size: 80%;">And a corresponding example using <b>scikit-learn</b></a></li>
<!-- navigation toc: --> <li><a href="._week39-bs041.html#___sec40" style="font-size: 80%;">Gradient descent and Ridge</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs042.html#___sec41" style="font-size: 80%;">Program example for gradient descent with Ridge Regression</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs043.html#___sec42" style="font-size: 80%;">Using gradient descent methods, limitations</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs044.html#___sec43" style="font-size: 80%;">Friday September 25</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs045.html#___sec44" style="font-size: 80%;">Stochastic Gradient Descent</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs046.html#___sec45" style="font-size: 80%;">Computation of gradients</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs047.html#___sec46" style="font-size: 80%;">SGD example</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs048.html#___sec47" style="font-size: 80%;">The gradient step</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs049.html#___sec48" style="font-size: 80%;">Simple example code</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs050.html#___sec49" style="font-size: 80%;">When do we stop?</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs051.html#___sec50" style="font-size: 80%;">Slightly different approach</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs052.html#___sec51" style="font-size: 80%;">Program for stochastic gradient</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs053.html#___sec52" style="font-size: 80%;">Momentum based GD</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs054.html#___sec53" style="font-size: 80%;">More on momentum based approaches</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs055.html#___sec54" style="font-size: 80%;">Momentum parameter</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs056.html#___sec55" style="font-size: 80%;">Second moment of the gradient</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs057.html#___sec56" style="font-size: 80%;">RMS prop</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs058.html#___sec57" style="font-size: 80%;">ADAM optimizer</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs059.html#___sec58" style="font-size: 80%;">Practical tips</a></li>
<!-- navigation toc: --> <li><a href="#___sec59" style="font-size: 80%;">Automatic differentiation</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs061.html#___sec60" style="font-size: 80%;">Using autograd</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs062.html#___sec61" style="font-size: 80%;">Autograd with more complicated functions</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs063.html#___sec62" style="font-size: 80%;">More complicated functions using the elements of their arguments directly</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs064.html#___sec63" style="font-size: 80%;">Functions using mathematical functions from Numpy</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs065.html#___sec64" style="font-size: 80%;">More autograd</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs066.html#___sec65" style="font-size: 80%;">And with loops</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs067.html#___sec66" style="font-size: 80%;">Using recursion</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs068.html#___sec67" style="font-size: 80%;">Unsupported functions</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs069.html#___sec68" style="font-size: 80%;">The syntax a.dot(b) when finding the dot product</a></li>
<!-- navigation toc: --> <li><a href="._week39-bs070.html#___sec69" style="font-size: 80%;">Recommended to avoid</a></li>
</ul>
</li>
</ul>
</div>
</div>
</div> <!-- end of navigation bar -->
<div class="container">
<p>&nbsp;</p><p>&nbsp;</p><p>&nbsp;</p> <!-- add vertical space -->
<a name="part0060"></a>
<!-- !split -->
<h2 id="___sec59" class="anchor">Automatic differentiation </h2>
<p>
<a href="https://en.wikipedia.org/wiki/Automatic_differentiation" target="_self">Automatic differentiation (AD)</a>,
also called algorithmic
differentiation or computational differentiation,is a set of
techniques to numerically evaluate the derivative of a function
specified by a computer program. AD exploits the fact that every
computer program, no matter how complicated, executes a sequence of
elementary arithmetic operations (addition, subtraction,
multiplication, division, etc.) and elementary functions (exp, log,
sin, cos, etc.). By applying the chain rule repeatedly to these
operations, derivatives of arbitrary order can be computed
automatically, accurately to working precision, and using at most a
small constant factor more arithmetic operations than the original
program.
<p>
Automatic differentiation is neither:
<ul>
<li> Symbolic differentiation, nor</li>
<li> Numerical differentiation (the method of finite differences).</li>
</ul>
Symbolic differentiation can lead to inefficient code and faces the
difficulty of converting a computer program into a single expression,
while numerical differentiation can introduce round-off errors in the
discretization process and cancellation
<p>
Python has tools for so-called <b>automatic differentiation</b>.
Consider the following example
$$
f(x) = \sin\left(2\pi x + x^2\right)
$$
which has the following derivative
$$
f'(x) = \cos\left(2\pi x + x^2\right)\left(2\pi + 2x\right)
$$
Using <b>autograd</b> we have
<p>
<!-- code=python (!bc pycod) typeset with pygments style "default" -->
<div class="highlight" style="background: #f8f8f8"><pre style="line-height: 125%"><span></span><span style="color: #008000; font-weight: bold">import</span> <span style="color: #0000FF; font-weight: bold">autograd.numpy</span> <span style="color: #008000; font-weight: bold">as</span> <span style="color: #0000FF; font-weight: bold">np</span>
<span style="color: #408080; font-style: italic"># To do elementwise differentiation:</span>
<span style="color: #008000; font-weight: bold">from</span> <span style="color: #0000FF; font-weight: bold">autograd</span> <span style="color: #008000; font-weight: bold">import</span> elementwise_grad <span style="color: #008000; font-weight: bold">as</span> egrad
<span style="color: #408080; font-style: italic"># To plot:</span>
<span style="color: #008000; font-weight: bold">import</span> <span style="color: #0000FF; font-weight: bold">matplotlib.pyplot</span> <span style="color: #008000; font-weight: bold">as</span> <span style="color: #0000FF; font-weight: bold">plt</span>
<span style="color: #008000; font-weight: bold">def</span> <span style="color: #0000FF">f</span>(x):
<span style="color: #008000; font-weight: bold">return</span> np<span style="color: #666666">.</span>sin(<span style="color: #666666">2*</span>np<span style="color: #666666">.</span>pi<span style="color: #666666">*</span>x <span style="color: #666666">+</span> x<span style="color: #666666">**2</span>)
<span style="color: #008000; font-weight: bold">def</span> <span style="color: #0000FF">f_grad_analytic</span>(x):
<span style="color: #008000; font-weight: bold">return</span> np<span style="color: #666666">.</span>cos(<span style="color: #666666">2*</span>np<span style="color: #666666">.</span>pi<span style="color: #666666">*</span>x <span style="color: #666666">+</span> x<span style="color: #666666">**2</span>)<span style="color: #666666">*</span>(<span style="color: #666666">2*</span>np<span style="color: #666666">.</span>pi <span style="color: #666666">+</span> <span style="color: #666666">2*</span>x)
<span style="color: #408080; font-style: italic"># Do the comparison:</span>
x <span style="color: #666666">=</span> np<span style="color: #666666">.</span>linspace(<span style="color: #666666">0</span>,<span style="color: #666666">1</span>,<span style="color: #666666">1000</span>)
f_grad <span style="color: #666666">=</span> egrad(f)
computed <span style="color: #666666">=</span> f_grad(x)
analytic <span style="color: #666666">=</span> f_grad_analytic(x)
plt<span style="color: #666666">.</span>title(<span style="color: #BA2121">&#39;Derivative computed from Autograd compared with the analytical derivative&#39;</span>)
plt<span style="color: #666666">.</span>plot(x,computed,label<span style="color: #666666">=</span><span style="color: #BA2121">&#39;autograd&#39;</span>)
plt<span style="color: #666666">.</span>plot(x,analytic,label<span style="color: #666666">=</span><span style="color: #BA2121">&#39;analytic&#39;</span>)
plt<span style="color: #666666">.</span>xlabel(<span style="color: #BA2121">&#39;x&#39;</span>)
plt<span style="color: #666666">.</span>ylabel(<span style="color: #BA2121">&#39;y&#39;</span>)
plt<span style="color: #666666">.</span>legend()
plt<span style="color: #666666">.</span>show()
<span style="color: #008000">print</span>(<span style="color: #BA2121">&quot;The max absolute difference is: </span><span style="color: #BB6688; font-weight: bold">%g</span><span style="color: #BA2121">&quot;</span><span style="color: #666666">%</span>(np<span style="color: #666666">.</span>max(np<span style="color: #666666">.</span>abs(computed <span style="color: #666666">-</span> analytic))))
</pre></div>
<p>
<p>
<!-- navigation buttons at the bottom of the page -->
<ul class="pagination">
<li><a href="._week39-bs059.html">&laquo;</a></li>
<li><a href="._week39-bs000.html">1</a></li>
<li><a href="">...</a></li>
<li><a href="._week39-bs052.html">53</a></li>
<li><a href="._week39-bs053.html">54</a></li>
<li><a href="._week39-bs054.html">55</a></li>
<li><a href="._week39-bs055.html">56</a></li>
<li><a href="._week39-bs056.html">57</a></li>
<li><a href="._week39-bs057.html">58</a></li>
<li><a href="._week39-bs058.html">59</a></li>
<li><a href="._week39-bs059.html">60</a></li>
<li class="active"><a href="._week39-bs060.html">61</a></li>
<li><a href="._week39-bs061.html">62</a></li>
<li><a href="._week39-bs062.html">63</a></li>
<li><a href="._week39-bs063.html">64</a></li>
<li><a href="._week39-bs064.html">65</a></li>
<li><a href="._week39-bs065.html">66</a></li>
<li><a href="._week39-bs066.html">67</a></li>
<li><a href="._week39-bs067.html">68</a></li>
<li><a href="._week39-bs068.html">69</a></li>
<li><a href="._week39-bs069.html">70</a></li>
<li><a href="">...</a></li>
<li><a href="._week39-bs070.html">71</a></li>
<li><a href="._week39-bs061.html">&raquo;</a></li>
</ul>
<!-- ------------------- end of main content --------------- -->
</div> <!-- end container -->
<!-- include javascript, jQuery *first* -->
<script src="https://ajax.googleapis.com/ajax/libs/jquery/1.10.2/jquery.min.js"></script>
<script src="https://netdna.bootstrapcdn.com/bootstrap/3.0.0/js/bootstrap.min.js"></script>
<!-- Bootstrap footer
<footer>
<a href="http://..."><img width="250" align=right src="http://..."></a>
</footer>
-->
<center style="font-size:80%">
<!-- copyright only on the titlepage -->
</center>
</body>
</html>