Machine Learning, On-line Algorithms, and Optimization

Avrim Blum

CCR-0105488


Summary (abbreviated):
The aim of this work is to develop new connections between fundamental problems in machine learning, online algorithms, and optimization. These connections will enable the production of new techniques and algorithms that will advance all three areas. Examples of work done under this grant include (1) using simple learning algorithms to solve basic optimization problems, (2) using online learning techniques to design adaptive data-structures and other online algorithms with improved guarantees, (3) graph-theoretic clustering problems motivated by machine learning, (4) use of online learning for ecommerce applications, and (5) connecting the machine learning problem of how to combine labeled and unlabeled data with combinatorial problems of designing good graph separators. In addition to developing connections between these areas, we are also interested in a number of basic problems in each area separately.

Survey talk (on connections between learning and online algorithms):
Online Learning Tools for Online Algorithms

Previous award:
CCR-9732707

Progress: 2001-2002


Progress: 2002-2003

Progress: 2003-2004


This material is based upon work supported by the National Science Foundation under Grant No. CCR-0105488. Any opinions, findings and conclusions or recomendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).

Avrim Blum
Last modified: Wed Jul 20 13:10:21 EDT 2005