Approximation of optimal two-dimensional association rules for categorical attributes using semidefinite programming

Katsuki Fujisawa, Yukinobu Hamuro, Naoki Katoh, Takeshi Tokuyama, Katsutoshi Yada

Research output: Chapter in Book/Report/Conference proceedingConference contribution

9 Citations (Scopus)

Abstract

We consider the problem of finding two-dimensional association rules for categorical attributes. Suppose we have two conditional attributes A and B both of whose domains are categorical, and one binary target attribute whose domain is {“positive”, “negative”}. We want to split the Cartesian product of domains of A and B into two subsets so that a certain objective function is optimized, i.e., we want to find a good segmentation of the domains of A and B. We consider in this paper the objective function that maximizes the confidence under the constraint of the upper bound of the support size. We first prove that the problem is NP-hard, and then propose an approximation algorithm based on semidefinite programming. In order to evaluate the effectiveness and efficiency of the proposed algorithm, we carry out computational ex- periments for problem instances generated by real sales data consisting of attributes whose domain size is a few hundreds at maximum. Approxi- mation ratios of the solutions obtained measured by comparing solutions for semidefinite programming relaxation range from 76% to 95%. It is observed that the performance of generated association rules are signifi- cantly superior to that of one-dimensional rules.

Original languageEnglish
Title of host publicationDiscovery Science - 2nd International Conference, DS 1999, Proceedings
EditorsSetsuo Arikawa, Koichi Furukawa
PublisherSpringer-Verlag
Pages148-159
Number of pages12
ISBN (Print)354066713X, 9783540667131
DOIs
Publication statusPublished - 1999 Jan 1
Event2nd International Conference on Discovery Science, DS 1999 - Tokyo, Japan
Duration: 1999 Dec 61999 Dec 8

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1721
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other2nd International Conference on Discovery Science, DS 1999
CountryJapan
CityTokyo
Period99/12/699/12/8

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Approximation of optimal two-dimensional association rules for categorical attributes using semidefinite programming'. Together they form a unique fingerprint.

  • Cite this

    Fujisawa, K., Hamuro, Y., Katoh, N., Tokuyama, T., & Yada, K. (1999). Approximation of optimal two-dimensional association rules for categorical attributes using semidefinite programming. In S. Arikawa, & K. Furukawa (Eds.), Discovery Science - 2nd International Conference, DS 1999, Proceedings (pp. 148-159). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1721). Springer-Verlag. https://doi.org/10.1007/3-540-46846-3_14