MILP approach to pattern generation in logical analysis of data

Hong Seo Ryoo, In Yong Jang

Research output: Contribution to journalArticle

45 Citations (Scopus)

Abstract

Pattern generation methods for the Logical Analysis of Data (LAD) have been term-enumerative in nature. In this paper, we present a Mixed 0-1 Integer and Linear Programming (MILP) approach that can identify LAD patterns that are optimal with respect to various previously studied and new pattern selection preferences. Via art of formulation, the MILP-based method can generate optimal patterns that also satisfy user-specified requirements on prevalence, homogeneity and complexity. Considering that MILP problems with hundreds of 0-1 variables are easily solved nowadays, the proposed method presents an efficient way of generating useful patterns for LAD. With extensive experiments on benchmark datasets, we demonstrate the utility of the MILP-based pattern generation.

Original languageEnglish
Pages (from-to)749-761
Number of pages13
JournalDiscrete Applied Mathematics
Volume157
Issue number4
DOIs
Publication statusPublished - 2009 Feb 28

Fingerprint

0-1 Integer Programming
Mixed Integer Programming
Integer programming
Linear programming
Homogeneity
Benchmark
Formulation
Experiments
Requirements
Term
Demonstrate
Experiment

Keywords

  • Combinatorial optimization
  • Logical analysis of data
  • Mixed 0-1 integer and linear programming
  • Pattern
  • Supervised machine learning

ASJC Scopus subject areas

  • Applied Mathematics
  • Discrete Mathematics and Combinatorics

Cite this

MILP approach to pattern generation in logical analysis of data. / Ryoo, Hong Seo; Jang, In Yong.

In: Discrete Applied Mathematics, Vol. 157, No. 4, 28.02.2009, p. 749-761.

Research output: Contribution to journalArticle

@article{d8f28451842a4169b2651fa27bbd8810,
title = "MILP approach to pattern generation in logical analysis of data",
abstract = "Pattern generation methods for the Logical Analysis of Data (LAD) have been term-enumerative in nature. In this paper, we present a Mixed 0-1 Integer and Linear Programming (MILP) approach that can identify LAD patterns that are optimal with respect to various previously studied and new pattern selection preferences. Via art of formulation, the MILP-based method can generate optimal patterns that also satisfy user-specified requirements on prevalence, homogeneity and complexity. Considering that MILP problems with hundreds of 0-1 variables are easily solved nowadays, the proposed method presents an efficient way of generating useful patterns for LAD. With extensive experiments on benchmark datasets, we demonstrate the utility of the MILP-based pattern generation.",
keywords = "Combinatorial optimization, Logical analysis of data, Mixed 0-1 integer and linear programming, Pattern, Supervised machine learning",
author = "Ryoo, {Hong Seo} and Jang, {In Yong}",
year = "2009",
month = "2",
day = "28",
doi = "10.1016/j.dam.2008.07.005",
language = "English",
volume = "157",
pages = "749--761",
journal = "Discrete Applied Mathematics",
issn = "0166-218X",
publisher = "Elsevier",
number = "4",

}

TY - JOUR

T1 - MILP approach to pattern generation in logical analysis of data

AU - Ryoo, Hong Seo

AU - Jang, In Yong

PY - 2009/2/28

Y1 - 2009/2/28

N2 - Pattern generation methods for the Logical Analysis of Data (LAD) have been term-enumerative in nature. In this paper, we present a Mixed 0-1 Integer and Linear Programming (MILP) approach that can identify LAD patterns that are optimal with respect to various previously studied and new pattern selection preferences. Via art of formulation, the MILP-based method can generate optimal patterns that also satisfy user-specified requirements on prevalence, homogeneity and complexity. Considering that MILP problems with hundreds of 0-1 variables are easily solved nowadays, the proposed method presents an efficient way of generating useful patterns for LAD. With extensive experiments on benchmark datasets, we demonstrate the utility of the MILP-based pattern generation.

AB - Pattern generation methods for the Logical Analysis of Data (LAD) have been term-enumerative in nature. In this paper, we present a Mixed 0-1 Integer and Linear Programming (MILP) approach that can identify LAD patterns that are optimal with respect to various previously studied and new pattern selection preferences. Via art of formulation, the MILP-based method can generate optimal patterns that also satisfy user-specified requirements on prevalence, homogeneity and complexity. Considering that MILP problems with hundreds of 0-1 variables are easily solved nowadays, the proposed method presents an efficient way of generating useful patterns for LAD. With extensive experiments on benchmark datasets, we demonstrate the utility of the MILP-based pattern generation.

KW - Combinatorial optimization

KW - Logical analysis of data

KW - Mixed 0-1 integer and linear programming

KW - Pattern

KW - Supervised machine learning

UR - http://www.scopus.com/inward/record.url?scp=59949094699&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=59949094699&partnerID=8YFLogxK

U2 - 10.1016/j.dam.2008.07.005

DO - 10.1016/j.dam.2008.07.005

M3 - Article

VL - 157

SP - 749

EP - 761

JO - Discrete Applied Mathematics

JF - Discrete Applied Mathematics

SN - 0166-218X

IS - 4

ER -