Information-theoretic framework for optimization with application to supervised learning

David Miller, Ajit Rao, Kenneth Rose, Allen Gersho

Research output: Contribution to conferencePaperpeer-review

2 Scopus citations

Abstract

A unified approach is developed for hard optimization problems involving data association, i.e. the assignment of elements viewed as 'data' to one of a set classes so as to minimize the resulting cost. The diverse problems which fit this description include data clustering, statistical classifier design to minimize probability of error, piece-wise regression structure vector quantization, as well as optimization in graph theory. Whereas standard descent-based methods are susceptible to finding poor local optima of the cost, the suggested approach provides some potential for avoiding local optima, yet without the computational complexity of stochastic annealing.

Original languageEnglish (US)
Number of pages1
StatePublished - Jan 1 1995
EventProceedings of the 1995 IEEE International Symposium on Information Theory - Whistler, BC, Can
Duration: Sep 17 1995Sep 22 1995

Other

OtherProceedings of the 1995 IEEE International Symposium on Information Theory
CityWhistler, BC, Can
Period9/17/959/22/95

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Information Systems
  • Modeling and Simulation
  • Applied Mathematics

Fingerprint Dive into the research topics of 'Information-theoretic framework for optimization with application to supervised learning'. Together they form a unique fingerprint.

Cite this