Class Times: Monday and Wednesday 10:30-12:00 Units: 3-0-9 H,G Location: E25-202 Instructors: Tomaso Poggio, Sayan Mukherjee, Ryan Rifkin, Alex Rakhlin
Office Hours: By appointment Email Contact : 9.520@mit.edu Previous Classes: SPRING 03 Course description
Focuses on the problem of supervised learning from the perspective of modern statistical learning theory starting with the problem of multivariate function approximation from sparse data. Develops basic tools such as Regularization including Support Vector Machines. Derives generalization bounds using both stability and covering number conditions. Describes applications in several areas, such as computer vision, computer graphics, text classification and bioinformatics. Several final projects -- some of which may be seeds for theses -- and hands-on applications and exercises are planned, given the rapidly increasing practical use of the techniques described in the subject.Prerequisites
18.02, 9.641, 6.893 or permission of instructor. In practice, a substantial level of mathematical maturity is necessary. Familiarity with probability and functional analysis will be very helpful. We try to keep the mathematical prerequisites to a minimum, but we will introduce complicated material at a fast pace.Grading
There will be two problem sets, a Matlab assignment, and a final project. To receive credit, you must attend regularly, and put in effort on all problem sets and the project.
Problem sets
Problem set #1: PS, PDF (updated 3/8/2004)
Problem set #2: PS, PDF
Projects
Project ideas: PS, PDF
Syllabus
Follow the link for each class to find a detailed description, suggested readings, and class slides. Some of the later classes may be subject to reordering or rescheduling.
Date Title Instructor(s) Class 01 Wed 04 Feb The Course at a Glance TP Class 02 Mon 9 Feb The Learning Problem in Perspective TP Class 03 Wed 11 Feb Regularization and Reproducing Kernel Hilbert Spaces TP,SM Class 04 Tue 17 Feb Regression and Least-Squares Classification RR Class 05 Wed 18 Feb Support Vector Machines for Classification RR Class 06 Mon 23 Feb Generalization Bounds, Intro to Stability AR Class 07 Wed 26 Feb Stability of Tikhonov Regularization AR Class 08 Mon 01 Mar Consistency and Uniform Convergence Over Function Classes AR Class 09 Wed 03 Mar Necessary and sufficient conditions for Uniform Convergence SM Class 10 Mon 8 Mar Stability and Glivenko-Cantelli Classes TP Class 11 Wed 10 Mar Multiclass Classification RR Class 12 Mon 15 Mar Computer Vision, Object Detection LW,SB Class 13 Wed 17 Mar Loose ends, Project discussions TP,SM,RR,AR SPRING BREAK Class 14 Mon 29 Mar Boosting and Bagging SM Class 15 Wed 31 Mar Text JR Class 16 Mon 05 Apr Approximation Theory FG Class 17 Wed 07 Apr Symmetrization, Rademacher Averages AR Class 18 Mon 12 Apr Regularization Networks TP Class 19 Wed 14 Apr Morphable Models for Video TE Class 20 Wed 21 Apr Leave-one-out approximations SM Class 21 Mon 26 Apr Computational biology GY,SM Class 22 Wed 28 Apr RKHS, Mercer Thm, Unbounded Domains, Frames and Wavelets TP,SM Class 23 Mon 03 May Unsupervised Learning, Learning with Partially Labeled Data,
Active Learning, Learning ManifoldsMB Class 24 Wed 05 May Bayesian Interpretations TP,SM Class 25 Mon 10 May Project Presentations Class 26 Wed 12 May Project Presentations
Math Camp 1 Mon 9 Feb Analysis and basic probability theory SM Math Camp 2 Tue 17 Feb More analysis and probability theory (Updated!) TP,AR Reading List
There is no textbook for this course. All the required information will be presented in the slides associated with each class. The books listed below are useful general reference reading, especially from the theoretical viewpoint. A list of suggested readings will also be provided separately for each class.
- V. N. Vapnik. The Nature of Statistical Learning Theory. Springer, 1995.
- V. N. Vapnik. Statistical Learning Theory. Wiley, 1998.
- L. Devroye, L. Gyorfi, and G. Lugosi. A Probabilistic Theory of Pattern Recognition. Springer, 1997.
- N. Cristianini and J. Shawe-Taylor. Introduction To Support Vector Machines. Cambridge, 2000.
- T. Evgeniou and M. Pontil and T. Poggio. Regularization Networks and Support Vector Machines. Advances in Computational Mathematics, 2000.
- F. Cucker and S. Smale. On The Mathematical Foundations of Learning. Bulletin of the American Mathematical Society, 2002.
- T. Poggio and S. Smale. The Mathematics of Learning: Dealing with Data. Notices of the AMS, 2003
- S. Mukherjee, P. Niyogi, T. Poggio and R. Rifkin. Statistical Learning: Stability is Sufficient for Generalization and Necessary and Sufficient for Consistency of Empirical Risk Minimization. AI/CBCL Memo, 2002 (revised 2003).
- Poggio, T., R. Rifkin, S. Mukherjee and P. Niyogi. General Conditions for Predictivity in Learning Theory, Nature, Vol. 428, 419-422, 2004 (see also Past Performance and Future Results).