"It is by logic we prove, but by intuition that we discover."
                                                                         -- Henri Poincare
Graphical Models and Applications

Course Objective:

This course aims to provide an introduction to the area of probabilistic models based on graphs, and meanwhile to offer an overall view on examining the recent trends of using graphical models in various application domains such as machine learning, sensor network, bioinformatics, signal/multimedia processing, and computer vision, etc. Through the study of key models, e.g., Bayesian networks, state space models, Markov random fields, and conditional random fields, we will investigate and test several popular algorithms related to exact or approximate inference, including elimination algorithm, junction tree, sum-product and max-product algorithms, loopy belief propagation, and variational methods. Students taking the course are required to submit a final term project that emphasizes an appropriate use of graphical model techniques on their choice of application.

Course Topics:

  • Basic probability theory and graph theory
  • Representations of graphical models
  • Learning from data
  • Methods for exact inference
  • Methods for approximate inference
  • Applications

Course Information:

Instructor: Hwann-Tzong Chen
TA: Tzu-Wei Huang
Time: T6R7R8
Location: Delta 102