Improved rule discovery performance on uncertainty

Mehmet R. Tolun, Hayri Sever, Mahmut Uludag

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

1 Scopus citations

Abstract

In this paper we describe the improved version of a novel rule induction algorithm, namely ILA. We first outline the basic algorithm, and then present how the algorithm is enhanced using the new evaluation metric that handles uncertainty in a given data set. In addition to having a faster induction than the original one, we believe that our contribution comes into picture with a new metric that allows users to define their preferences through a penalty factor. We use this penalty factor to tackle with over-fitting bias, which is inherently found in a great many of inductive algorithms. We compare the improved algorithm ILA-2 to a variety of induction algorithms, including ID3, OCI, C4.5, CN2, and ILA. According to our preliminary experimental work, the algorithm appears to be comparable to the well-known algorithms such as CN2 and C4.5 in terms of accuracy and size.

Original languageEnglish (US)
Title of host publicationResearch and Development in Knowledge Discovery and Data Mining - 2nd Pacific-Asia Conference, PAKDD 1998, Proceedings
EditorsXindong Wu, Ramamohanarao Kotagiri, Kevin B. Korb
PublisherSpringer Verlag
Pages310-321
Number of pages12
ISBN (Print)3540643834, 9783540643838
DOIs
StatePublished - Jan 1 1998
Event2nd Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 1998 - Melbourne, Australia
Duration: Apr 15 1998Apr 17 1998

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1394
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other2nd Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 1998
CountryAustralia
CityMelbourne
Period04/15/9804/17/98

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Improved rule discovery performance on uncertainty'. Together they form a unique fingerprint.

Cite this