Instructor:  Milos Hauskrecht
Computer Science Department
MIB 318
phone: x4-8845
e-mail: milos@cs.pitt.edu
office hours: by appointment

TA:  Mathew Bell
Computer Science Department
MIB 317
phone: x4-8842
e-mail: mbell@cs.pitt.edu
office hours: Monday, Tuesday 10 - 11:30 am

Term projects reports are due on Wednesday, April 24, 2002 at 4pm.

Project presentations will be held on the same day, April 24, 2002, starting at 4pm (for about 3 1/2 hours). The presentations should be:

• 15 minutes long (12 minutes presentation + 3 minutes questions);
• Describe the main parts of the project: data, learning problem, learning methods, results obtained, comparison and evaluation of results. )

Course description
Lectures
Other course information (readings, grading, homeworks)
Term projects (Project ideas and other resources )
Matlab

The goal of the field of machine learning is to build computer systems that learn from experience and that are capable of adapting to their environments. Learning techniques and methods developed by researchers in this field have been successfully applied to a variety of learning tasks in a broad range of areas, including, for example, text classification, gene discovery, financial forecasting, credit card fraud detection, collaborative filtering, design of adaptive web agents and others.

This introductory machine learning course will give an overview of many techniques and algorithms in machine learning, beginning with topics such as simple concept learning and ending up with more recent topics such as boosting, support vector machines, and reinforcement learning. The  objective of the course  is  not only to  present  the modern machine learning methods but also to give the basic intutions behind the methods as well as, a more formal understanding of how and why they work.

Topics (subject to minor changes)

• Regression. Loss function. Least-squares fit. Parameter estimation. Statistical view on the regression. Gradient descent. On-line learning.
• Classification. Discriminant functions. Logistic regression. Gradient descent. Multiple classes (softmax). Generative classification models. Parameter estimation.
• Neural networks. Backpropagation.
• Support vector machines. Classification. Max margin hyperplanes. Kernel functions.
• Density estimation. Parametric methods.
• Bayesian networks. Independence structure. Inference. Parameter learning. Structure learning.
• Density estimation with hidden variables and missing values. Expectation-maximization algorithm.
• Clustering. Mixture of Gaussians. k-means. Hierachical methods.
• Non-parametric density estimation. Parzen windows. Nearest neighbor.
• Dimensionality reduction. Feature extraction. Mutual information measure. PCA. Reductions based on clustering.
• Ensamble methods. Mixture of experts. Bagging, boosting.
• Hidden Markov models. Forward, backward algorithm. Viterbi algorithm. Baum-Welch algorithm.
• Reinforcement learning. Learning to act. Markov decision processes. Reinforcement learning with delayed rewards.
• Concept learning. Version spaces. PAC learning. VC dimension.

Undergraduate level knowledge of matrices and linear algebra (CS 0280), probability (CS 1151), programming (CS 1501) or equivalent, or the permission of the instructor.

Readings will be distributed electronically through the course web page

Primary readings:
M. Jordan, C. Bishop. Introduction to graphical models. Local Upitt access only!!

Some readings for future classes (in electronic form)

Recommended books:
Chris Bishop. Neural networks for Pattern Recognition. Oxford University Press, 1995.
R.O. Duda, P.E. Hart, D.G. Stork.  Pattern Classification. Second edition. John Wiley and Sons, 2000.
T. Mitchell.  Machine Learning. Mc Graw Hill, 1997.

Other related books you may find useful:
Han, Kamber. Data mining. Concepts and Techniques. Morgan Kauffman, 2001.

The course grade will be determined roughly as follows :

• Lectures                                10%
• Homework assignments         40%
• Final project and Exam(s)      50 %

The homework assignments will have mostly a character of projects and will require you to implement some of the learning algorithms covered during the semester and apply them. No collaboration on homeworks is allowed. To implement homework assignments you may use a programming language and graphic tools of your choice. However, students are highly enouraged to use Matlab.
 

Term project is due at the end of the semester and accounts for about 40% of your grade. You can choose your own problem topic. You will be asked to write a short proposal for the purpose of approval and feedback. The project must have a distinctive and non-trivial learning or adaptive component. In general, a project may consist of a replication of previously published results, design of new learning methods and their testing, or application of machine learning to a domain or a problem of your interest.

Matlab is a mathematical tool for numerical computation and manipulation, with excellent graphing capabilities. It provides a great deal of support and capabilities for things you will need to run Machine Learning experiments. Upitt has a number of Matlab licences running on both unix and windows platforms. Click here to find out how to access Matlab at Upitt.

Files used during the tutorial on 01/16/02.

Other Matlab resources on the web:

Online MATLAB  documentation
Online Mathworks documentation including MATLAB toolboxes