Tuesdays & Thursday 3:05 - 4:25PM in Klaus 1456

Henrik I Christensen
Office: CCB216
Hours: Tues 2-3
TA: Siddharth Choudhary
Office: 273A
Office Hours: Fri 3-4pm
TA: Steven Hickson
Office: CCB
Office Hours: Wed 11am-12pm
TA: Ruffin White
Office Hours: Fri 2-3pm

Who is this for?

This is a graduate level for those interested in pattern recognition in general and for some elements as applied to computer vision. It is not going to be a comprehensive Machine Learning course.


Since this is a graduate course we are typically quite loose with prerequisites.
  • A good foundation of probability and linear algebra. This class will have more math in it than most Computer Science classes.
  • Any Machine Learning background will help. Though the course won’t technically presume ML as a background it will be much easier to grasp if you’ve seen things like graphical models or other inference structures.
  • A good working knowledge of Matlab or Python with Numpy. We will likely be doing things in Matlab in class. I am pretty sure that Octave will be OK though the lack of some plotting make may some figures harder to generate. Because this is my first time offering this class we will experiment.


The two texts that we will draw from are:
  • "The Elements of Statistical Learning: Data Mining, Inference, and Prediction. (2nd Ed)" by Hastie, Tibshirani, and Friedman (web site, Amazon). The book is available as a free PDF from the web site but it is also only $60 from Amazon which is relatively inexpensive for such a good text.
  • “Machine Learning: a Probabilistic Perspective” by Kevin Murphy. (web site, Amazon). THERE IS A LEGAL PDF in the GT Library (https://portal.library.gatech.edu/vufind/Record/1195667) but this is a best-seller (well, best selling ML book on Amazon) and is also relatively inexpensive at about $79.

We will also use a more basic text:
  • “Pattern Classification” by Duda, Hart and Stork (Amazon) This is a classic text (used to be just Duda and Hart) that I first learned PatRec in. In some sense it mostly predated “modern” Machine Learning, but it’s extremely accessible.

We will require readings from both Hastie and Murphy. In additon we will read a number of current day research papers to give you a sense of how the basic techniques are applied to real-world problems.


  • Web site: This site. Will have posted calendar/syllabus with posted slides, problem sets with data, other administrative stuff.
  • T-square: The usual stuff. There is a web page under resources that points to this class web page.
  • Slides: PDF slides will be posted by linking to the calendar. As these will be collaborative lectures (more on that in class) they will probably not be posted before class.
  • Piazza: You will all receive an invitation to Piazza for CS7616. If not, send us email. Rather than emailing questions to the teaching staff (ie me), I encourage you to post your questions on Piazza. Find our class page at:
  • Matlab access: if you don’t know how to get MATLAB access, first ask a friend. Then come see the TAs or me. I am convinced that Octave will be adequate. Python access: some of you may use Python and Numpy. We’ll try to give you support for that too.


The grade will be based upon a small number of projects some of which can be done in groups no larger than two. The intent is to have three projects where everyone in the class uses the same data set and a variety of algorithms, whereas for the “final” project you will need to propose your own pattern recognition problem/data set. Also, everyone will be responsible for presenting some Pattern Recognition material, based on more recent methods in research papers. The presentations will be in smaller groups of 15-20 students to enable more discussion.
  • Communal Projects: 45%
  • Final project: 25%
  • Class presentation: 25%
  • Class participation: max 5% (ie it can only raise your grade)

Late Policy

Don't be late, it’s not fair to the TA. We will not accept late home works for full credit it unless you have (several day) prior approval from the professor for really extenuating circumstances. Submit what you have on time.

Honesty and Integrity Policy

Projects are to be done individually (or within your group) but you may collaborate at the white board level. You may help each other with algorithms and general computation, but your code must be your own.
Last modified: 1/1/2016