Convex and Distributed Optimization - WMM9MO15

Objectif(s)

This course is an introduction to stochastic and distributed algorithms.

Contenu(s)

Outline of the course:

• Introduction to convex optimization: concepts of convex analysis (duality, proximal operators), how to identify potential difficulties in optimization problems, specific difficulties of large-scale settings. Illustrations in supervised learning (classification and regression problems) and in operation research (decomposition methods).

• Stochastic algorithms (SGD and variance-reduced versions) and applications in learning

• Distributed algorithms (map-reduce framework, asynchronous algorithms, federated learning)

Prerequisite: fundamental knowledge on matrix analysis and optimisation (corresponding to the Refresher course given at the beginning of the semester) is required. Basic notions and terminology intersect the material of the course "Efficient methods in optimization".