[(Optimization for Machine Learning)] [Author: Suvrit Sra] published on

by Suvrit Sra

The interplay between optimization and machine learning is one of themost important developments in modern computational science. Optimizationformulations and methods are proving to be vital in designing algorithms to extractessential knowledge from huge volumes of data. Machine learning, however, is notsimply a consumer of optimization technology but a rapidly evolving field that isitself generating new optimization ideas. This book captures the state of the art ofthe interaction between optimization and machine learning in a way that isaccessible to researchers in both fields.Optimization approaches have enjoyedprominence in machine learning because of their wide applicability and attractivetheoretical properties. The increasing complexity, size, and variety of today'smachine learning models call for the reassessment of existing assumptions. This bookstarts the process of reassessment. It describes the resurgence in novel contexts ofestablished frameworks such as first-order methods, stochastic approximations, convex relaxations, interior-point methods, and proximal methods. It also devotesattention to newer themes such as regularized optimization, robust optimization, gradient and subgradient methods, splitting techniques, and second-order methods.Many of these techniques draw inspiration from other fields, including operationsresearch, theoretical computer science, and subfields of optimization. The book willenrich the ongoing cross-fertilization between the machine learning community andthese other fields, and within the broader optimization community.

Genres: Programming

Hardcover

First published September 30, 2011