Graphical Models for Optimization, Inference and Learning (Informal and Brief Tutorial)
When
5 p.m., Nov. 5, 2019
Graphical Models can be viewed as (a) universal programming language which allows to express and embed in Optimization, Inference and Learning (OIL) problems domain specific constraints, relations, bounds, limitations and symmetries between variables; (b) framework which generalizes multitude of methods in Data Science and Machine Learning. In this informal and brief tutorial I will explain how: (a) to set up a graphical model to account for an application specificity; (b) to pose OIL problems within the GM framework; (c) to resolve the GM problems, sometimes exactly but more often approximately.