Kalender

Course Introduction

Pattern Classification Overview. Features and feature distributions. Book Ch. 1.

Gaussian mixture model

Probabilistic Classification Theory

Review of Conditional Probability. General classifier design problem. Book Ch. 2, 3 - 3.2.

Probabilistic Classification Theory

Classifier decision criteria. Special case: Linear discriminant functions. Book Ch. 3 - 3.8. HMM sub-types. Book Ch. 5- 5.3.

Hidden Markov Model Algorithms

Forward Algorithm. Book Ch. 5.4.

HMM Algorithms

HMM Backward algoritm. Book Ch. 5.5.

Ch. 5

HMM continues

The backward algorithm and Viterbi algorithm, Book Ch. 5.5-5.6

Ch. 5

HMM Training

The Baum-Welch Algorithm, Book Ch. 6

Ch. 7

Expectation - Maximization 

Probabilistic model training, EM steps, theory, GMM training, Book Ch. 7.1 - 7.5 

Ch. 7

HMM Traning using EM, Book Ch. 7.6

Introduction to Bayesian Learning

Ch. 8

Bayesian learning

Procedure, Prior distribution, Conjugate distributions, Book Ch. 8 

Ch. 8

Approximate Bayesian Learning - Variational Inference, Book Ch. 9.1-9.3

After break: Review of the course and outlook (Saikat, Jalil and Frederik)

New location: OV. 10, Conference Room SIP (Floor 3) 

New location: OV. 10, Conference Room SIP (Floor 3)