STA 4273H (Fall 2012): Statistical Machine Learning

*** TUESDAY 2pm - 5pm in UC 175 ***

Instructor: Ruslan Salakhutdinov; email rsalakhu at utstat dot toronto dot edu

Lecture Times: Tuesday 2pm -- 5pm
Lecture Location: UC 175
First Lecture: Sep 11, 2012
Last Lecture: Dec 4, 2012
Office hours: Tuesdays 12-1pm

Prerequisite: Knowledge of statistical inference, probability theory, and linear algebra at the advanced undergraduate level, and some basic programming skills in R or Matlab. STA414/2104 is a plus, but is not required.

Marking Scheme
3 assignments worth 60%, one project worth 40%

Books :
Christopher M. Bishop (2006) Pattern Recognition and Machine Learning, Springer.
Trevor Hastie, Robert Tibshirani, Jerome Friedman (2009) The Elements of Statistical Learning
David MacKay (2003) Information Theory, Inference, and Learning Algorithms

Auditing
If you are not registered in the class, it is possible for you to audit it (sit in on the lectures), but only if you get the instructor's permission.

Course Description
This is an advanced graduate course, designed for Master's and Ph.D. level students, and will assume a reasonable degree of mathematical maturity. Specific topics to be covered include:

• Linear methods for regression/classification
• Model assessment and selection
• Graphical models, Bayesian networks, Markov random fields, conditional random fields
• Approximate variational inference, mean-field inference
• Basic sampling algorithms, Markov chain Monte Carlo, Gibbs sampling, and Metropolis-Hastings algorithm
• Mixture models and generalized mixture models
• Unsupervised learning, probabilistic PCA, factor analysis, independent component analysis.

STA 4273H (Fall 2012): Research Topics In Statistical Machine Learning || http://www.utstat.toronto.edu/~rsalakhu/sta4273_2012/