Personal:
  Program
  Achievements

Introduction:
  Problem
  Dist. Measure
  Approach
  Variants

Multivariate Gaussian:
  Definition
  Why Supp. Region?
  Support Region
  Identity
  Unequal Variances
  Off-diagonal
  Unconstrained

PCA:
  Math
  Eigen Params
  PCA Succeeds
  PCA Fails

LDA:
  Approach
  Math
  Opt. Criterion
  Class-dep
  LDA Vs PCA

Examples:
  Uniform Ellipse
  Parallel Ellipse
  Class-dep
  Max. Variance

Application:
  Overview
  Database
  Algorithm
  Evaluation
  Best Features
  Best System
  Why LDA Failed?

Conclusions:
  Summary
  References