CS 281B / Stat 241B, Spring 2006:

Statistical Learning Theory

Syllabus

Course description

This course will provide an introduction to the analysis of advanced statistical and computational methods for the modeling of complex, multivariate data. It will concentrate on the development of theoretical concepts to support such methods, and in particular the analysis of the statistical properties of prediction methods.
Prerequisites: CS281A/Stat241A, or advanced training in probability or statistics, at the level of Stat 205A or Stat 210A.

Outline:

  • Prediction problems
  • Kernel methods
    Reproducing kernel Hilbert spaces
    Representer theorem
    Cost functions and optimization
    Support vector machines
    Kernels
  • Ensemble methods
    Bagging
    Boosting
    Cost functions
    Optimization
    Classification versus regression
  • Model selection, regularization
  • Error estimates
    Concentration of measure
    Empirical processes
    Uniform convergence
    Complexity measures
    Empirical minimization
    Implications for linear methods


    • Back to course home page