Checking date: 19/05/2022

Course: 2022/2023

Inference methods in Bayesian Machine Learning
Study: Bachelor in Data Science and Engineering (350)

Coordinating teacher: MIGUEZ ARENAS, JOAQUIN

Department assigned to the subject: Department of Signal and Communications Theory

Type: Electives
ECTS Credits: 6.0 ECTS


Skills and learning outcomes
Description of contents: programme
In this course the student will advance in the study of inference methods for learning in probabilistic models. The objective of the course is to provide the student with an overview of the various approaches proposed to date in modern applications of Machine Learning. Each of the techniques will be illustrated on the basis of representative probabilistic models within the state of the art. PART I: SAMPLING METHODS 1. Basic Methods of Sampling. 2. MCMC methods: Metropolis-Hastings, Gibbs sampling. PART II: APPROXIMATE INFERENCE 3. Inference in graphic models using methods of dynamic programming and message passing. 4. Variational Inference and the Mean Field approximation. 5. Methods of Propagation of Hopes. PART III: SCALABLE METHODS TO LARGE DATASETS 6. Distributed and parallel MCMC 7. Stochastic Optimization in Variational Inference.
Assessment System
  • % end-of-term-examination 60
  • % of continuous assessment (assigments, laboratory, practicals...) 40

The course syllabus may change due academic events or other reasons.