Query by committee, linear separation and random walks

Ran Bachrach, Shai Fine, Eli Shamir

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

Recent works have shown the advantage of using Active Learning methods, such as the Query by Committee (QBC) algorithm, to various learning problems. This class of Algorithms requires an oracle with the ability to randomly select a consistent hypothesis according to some predefined distribution. When trying to implement such an oracle, for the linear separators family of hypotheses, various problems should be solved. The major problem is time-complexity, where the straight-forward Monte Carlo method takes exponential time. In this paper we address some of those problems and show how to convert them to the problems of sampling from convex bodies or approximating the volume of such bodies. We show that recent algorithms for approximating the volume of convex bodies and approximately uniformly sampling from convex bodies using random walks, can be used to solve this problem, and yield an eficient implementation for the QBC algorithm. This solution suggests a connection between random walks and certain properties known in machine learning such as ε-net and support vector machines. Working out this connection is left for future work.

Original languageEnglish
Title of host publicationComputational Learning Theory - 4th European Conference, EuroCOLT 1999, Proceedings
EditorsPaul Fischer, Hans Ulrich Simon
PublisherSpringer Verlag
Pages34-49
Number of pages16
ISBN (Print)3540657010, 9783540657019
DOIs
StatePublished - 1999
Event4th European Conference on Computational Learning Theory, EuroCOLT 1999 - Nordkirchen, Germany
Duration: 29 Mar 199931 Mar 1999

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1572
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference4th European Conference on Computational Learning Theory, EuroCOLT 1999
Country/TerritoryGermany
CityNordkirchen
Period29/03/9931/03/99

Bibliographical note

Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 1999.

Fingerprint

Dive into the research topics of 'Query by committee, linear separation and random walks'. Together they form a unique fingerprint.

Cite this