Master
2025/2026
Sampling and Generative Modeling
Type:
Compulsory course (Math of Machine Learning)
Delivered by:
Big Data and Information Retrieval School
Where:
Faculty of Computer Science
When:
1 year, 3 module
Open to:
students of one campus
Language:
English
ECTS credits:
3
Contact hours:
40
Course Syllabus
Abstract
In this course we present an introduction to different techniques of generating random variables from a given probability distribution. We focus on the setting when the distribution is known analytically, that is, on the sampling problem. At the same time, we mention connections with the generative modeling setup, when input data is represented by a fixed sample from the distribution of interest. In our course we cover some recent developments in MCMC algorithms, and consider other approaches, such as generative flow networks (GFlowNets) and diffusion samplers.