Skip to main navigation Skip to search Skip to main content

A new metric splitting criterion for decision trees

  • West Pomeranian University of Technology

Research output: Contribution to journalArticlepeer-review

Abstract

We examine a new approach to building decision tree by introducing a geometric splitting criterion, based on the properties of a family of metrics on the space of partitions of a finite set. This criterion can be adapted to the characteristics of the data sets and the needs of the users and yields decision trees that have smaller sizes and fewer leaves than the trees built with standard methods and have comparable or better accuracy.

Original languageEnglish
Pages (from-to)239-256
Number of pages18
JournalInternational Journal of Parallel, Emergent and Distributed Systems
Volume21
Issue number4
DOIs
StatePublished - Aug 1 2006

ASJC Scopus Subject Areas

  • Software
  • Computer Networks and Communications

Keywords

  • Decision tree
  • Generalized conditional entropy
  • Metric
  • Metric betweenness

Fingerprint

Dive into the research topics of 'A new metric splitting criterion for decision trees'. Together they form a unique fingerprint.

Cite this