CS 726: Advanced Machine Learning (Spring 2014)
Lecture Schedule Slot 9, Monday, Thursday: 3:30--5:00
Venue SIC 201, KR Building, CSE Department.
Instructor: Sunita Sarawagi
TA: Ahbirut Gupta
and Manasa Gootla
Instructor's office hours check here
Syllabus and week-wise calendar: Click here
Homeworks Click here
Previous years' exams
Course description
This course will be primarily about building supervised and
unsupervised learning models involving a very large number of
unknowns. The core techniques will be from two primary areas in
machine learning: probabilistic graphical models and learning in
structured output spaces. Topics include:
-
Representing directed and
undirected graphical models and understanding the relationship between
the trinity of conditional independencies, graph structure, and
factorization of the distribution.
- Exact and approximate inference algorithms with approaches spanning message passing, optimization, combinatorial algorithms, and sampling.
- Learning graph potentials and structure. Likelihood and max-margin methods for structured learning.
- Scaling issues with very large graphical model
- Applications from vision, text and web mining, health, and social media analysis.
Prerequisites
Foundation in machine learning (CS 725) or equivalent (example, CS 419).
Eligibility
The
course is open to CS MTechs, PhD, DD and BTech students. Students of
other departments should approach for permission only if they meet the
necessary pre-requisites. Third year BTech students need to take
prior permission from the instructor for enrolling in the course.
Credit/Audit Requirements
Approximate credit structure
- 25% Mid-semester exam
- 45% End semester exam
- 30% Homeworks + Quizzes
Audit students have to score more than 30% over all, including
assignments, quizzes and mid/end sem exam.
Reading List
Primary text book
Probabilistic Graphical Models: Principles and Techniques,
by Daphne Koller and Nir Friedman, MIT Press, 2009.
The course calendar will provide links
to other relevant papers and book chapters for specific topics.
Other relevant text books:
- R. G. Cowell, A. P. Dawid, S. L. Lauritzen and D. J.
Spiegelhalter. "Probabilistic Networks and Expert Systems".
Springer-Verlag. 1999.
-
M. I. Jordan (ed). "Learning in Graphical Models". MIT Press. 1998.
Collection of papers. These appear collated here.
-
J. Pearl. "Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference." Morgan Kaufmann. 1988.
- Graphical models by Lauritzen, Oxford science publications
- F. V. Jensen. "Bayesian Networks and Decision Graphs". Springer. 2001.
Survey articles