ml.html

<html>
  <head>
    <title>Machine Learning and Me</title>
    <link href="style.css" rel="stylesheet" type="text/css">
    
    <!-- Google Analytics Script -->
    <script type="text/javascript">

      var _gaq = _gaq || [];
      _gaq.push(['_setAccount', 'UA-16955664-1']);
      _gaq.push(['_trackPageview']);

      (function() {
      var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true;
      ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js';
      var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s);
      })();

    </script>
  </head>
  <body>
    <div id="topcontainer">
      <p id="toplinks"><a class="toplink" href="/index.html">Home</a><a class="toplink" href="/code.html">Code</a><a class="toplink" href="">Resume</a></p>
      <h2>Learning to do Machine Learning</h2>
      <p>There is so much data all around. Companies like google, facebook, amazon used data and machine learning(and information retrieval) to come up with various cool features that gave them unique competitive advantages. Netflix had a million dollar price competition to improve their prediction algorithms. Somehow motivated with all of this, I became excited with the domain of machine learning and started exploring.
I got a lot of guidance from following 2 resources:
<p><a href="http://pindancing.blogspot.com/2010/01/learning-about-machine-learniing.html">Ravi Mohan's Learning about Machine Learning post</a></p>
<p><a href="http://measuringmeasures.com/blog/2010/3/12/learning-about-machine-learning-2nd-ed.html">Bradford Cross's Learning about Machine Learning post</a></p>

<p>From above and a little peeking into <a href="http://research.microsoft.com/en-us/um/people/cmbishop/prml/">Pattern Recognition and Machine Learning</a> it became clear that I can not make any serious progress without actually having firm grounding in how to prove stuff, discrete maths, probability, statistics, calculus, linear algebra etc.</p>

<p>So I thought I will spend a year to cover all the prerequisites :). I don't know where all of this is going to take me but it sure is fun. So, Here is my ongoing progress...</p>

<div>
<ol>
<li>I worked through and completed pretty much every exercise in <a href="http://www.amazon.com/gp/product/0521675995/">How to prove it</a> to learn to think like mathematicians. All my progress with this book is noted in <a href="http://technotes-himanshu.blogspot.com/search/label/htpi">these posts</a> on my blog.</li>

<li>For discrete mathematics, I followed all the video lectures from <a href="http://www.archive.org/details/arsdigita_02_discrete_math">this course</a> and worked through pretty much every problem set. Again, <a href="http://technotes-himanshu.blogspot.com/search/label/cs-discrete-maths">these posts</a> contain all my progress.</li>

<li>For probability, I worked through <a href="http://technotes-himanshu.blogspot.com/search/label/probability">Introduction to Probability</a> and finished problem sets in <a href="http://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-041-probabilistic-systems-analysis-and-applied-probability-spring-2006/">introductory probability course at MIT</a>. Again, the progress is noted on <a href="http://technotes-himanshu.blogspot.com/search/label/probability">these posts</a>.</li>

<li>Just to get an overview of machine learning algorithms, I followed video lectures provided in <a href="http://technotes-himanshu.blogspot.com/search/label/ml">stanford machine learning course</a>. Also, tried to captured the high level learnings <a href="http://technotes-himanshu.blogspot.com/search/label/ml">here</a>.</li>

<li>To get accustomed with basic AI stuff, I implemented many algorithms(in Scala) from <a href="http://aima.cs.berkeley.edu/">Artificial Intelligence: A Modern Approach</a>. Here is the <a href="http://code.google.com/p/aima-scala/">code</a>.</li>

<li>For basic statistics, I worked through some chapters in <a href="http://www.amazon.com/gp/product/0534399428/">Mathematical Statistics and Data Analysis</a> and did the problem sets from <a href="http://ocw.mit.edu/courses/mathematics/18-443-statistics-for-applications-spring-2009/index.htm">MIT-Statistics for Applications course</a>. Again, the progress is <a href="http://technotes-himanshu.blogspot.com/search/label/statistics">here</a>. </li>
<li>Now a days, I am working my way through the overwhelming amount of maths in <a href="http://research.microsoft.com/en-us/um/people/cmbishop/prml/">Pattern Recognition and Machine Learning</a></li>
</ol>
</div>



    </div>
  </body>
</html>