Comments on Information and Distance Measures, Error Bounds, and Feature Subset Selection.

Abstract

A recent paper by L. Kanal provides a fairly complete list of information and distance measures and error bounds. This paper discusses the major problem areas with these measures, provides a more extensive list of measures including the newly proposed average conditional cubic entropy and suggests methods of approach for the unresolved problems. The topic of feature subset selection is also examined.

Document Details

Document Type
Technical Report
Publication Date
Jan 14, 1975
Accession Number
ADA005727

Entities

People

  • Chia‐Hung Chen

Organizations

  • University of Massachusetts Dartmouth

Tags

Fields of Study

  • Computer science

Readers

  • Graph Algorithms and Convex Optimization.
  • Theoretical Analysis.