Selecting structural patterns for classification

Wan Shiou Yang, San Yih Hwang, Jaideep Srivastava

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

Abstract

Many techniques have recently been proposed for discovering structural patterns. Using the discovered structural patterns as features for classification has shown success in some application domains. However, the efficiency and effectiveness of such a classification algorithm is often impeded by the huge number of structural patterns discovered by the associated structural pattern mining algorithm. In this paper, we focus on the feature selection problem of structural patterns. The goal is to develop a scheme that effectively selects a subset of structural patterns as the features for the following induction algorithm. We show how to make use of the downward closure property inherent in the structural patterns to design a novel feature selection algorithm. We also evaluate our algorithm by applying the real-world health insurance data for building a classification model to detect health care fraud and abuse. The experimental results show that a great extent of redundant features can be eliminated by our feature selection algorithm, resulting in both accuracy improvement and computation cost reduction.

Original languageEnglish
Title of host publicationProceedings of the Annual Hawaii International Conference on System Sciences
EditorsR.H. Spraque, Jr.
Pages55
Number of pages1
Publication statusPublished - 2005
Externally publishedYes
Event38th Annual Hawaii International Conference on System Sciences - Big Island, HI, United States
Duration: 3 Jan 20056 Jan 2005

Other

Other38th Annual Hawaii International Conference on System Sciences
CountryUnited States
CityBig Island, HI
Period3/1/056/1/05

Fingerprint

Feature extraction
Health insurance
Cost reduction
Health care

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Yang, W. S., Hwang, S. Y., & Srivastava, J. (2005). Selecting structural patterns for classification. In R. H. Spraque, Jr. (Ed.), Proceedings of the Annual Hawaii International Conference on System Sciences (pp. 55)

Selecting structural patterns for classification. / Yang, Wan Shiou; Hwang, San Yih; Srivastava, Jaideep.

Proceedings of the Annual Hawaii International Conference on System Sciences. ed. / R.H. Spraque, Jr. 2005. p. 55.

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

Yang, WS, Hwang, SY & Srivastava, J 2005, Selecting structural patterns for classification. in RH Spraque, Jr. (ed.), Proceedings of the Annual Hawaii International Conference on System Sciences. pp. 55, 38th Annual Hawaii International Conference on System Sciences, Big Island, HI, United States, 3/1/05.
Yang WS, Hwang SY, Srivastava J. Selecting structural patterns for classification. In Spraque, Jr. RH, editor, Proceedings of the Annual Hawaii International Conference on System Sciences. 2005. p. 55
Yang, Wan Shiou ; Hwang, San Yih ; Srivastava, Jaideep. / Selecting structural patterns for classification. Proceedings of the Annual Hawaii International Conference on System Sciences. editor / R.H. Spraque, Jr. 2005. pp. 55
@inproceedings{5a55618370654042bb81c53a5dd08512,
title = "Selecting structural patterns for classification",
abstract = "Many techniques have recently been proposed for discovering structural patterns. Using the discovered structural patterns as features for classification has shown success in some application domains. However, the efficiency and effectiveness of such a classification algorithm is often impeded by the huge number of structural patterns discovered by the associated structural pattern mining algorithm. In this paper, we focus on the feature selection problem of structural patterns. The goal is to develop a scheme that effectively selects a subset of structural patterns as the features for the following induction algorithm. We show how to make use of the downward closure property inherent in the structural patterns to design a novel feature selection algorithm. We also evaluate our algorithm by applying the real-world health insurance data for building a classification model to detect health care fraud and abuse. The experimental results show that a great extent of redundant features can be eliminated by our feature selection algorithm, resulting in both accuracy improvement and computation cost reduction.",
author = "Yang, {Wan Shiou} and Hwang, {San Yih} and Jaideep Srivastava",
year = "2005",
language = "English",
pages = "55",
editor = "{Spraque, Jr.}, R.H.",
booktitle = "Proceedings of the Annual Hawaii International Conference on System Sciences",

}

TY - GEN

T1 - Selecting structural patterns for classification

AU - Yang, Wan Shiou

AU - Hwang, San Yih

AU - Srivastava, Jaideep

PY - 2005

Y1 - 2005

N2 - Many techniques have recently been proposed for discovering structural patterns. Using the discovered structural patterns as features for classification has shown success in some application domains. However, the efficiency and effectiveness of such a classification algorithm is often impeded by the huge number of structural patterns discovered by the associated structural pattern mining algorithm. In this paper, we focus on the feature selection problem of structural patterns. The goal is to develop a scheme that effectively selects a subset of structural patterns as the features for the following induction algorithm. We show how to make use of the downward closure property inherent in the structural patterns to design a novel feature selection algorithm. We also evaluate our algorithm by applying the real-world health insurance data for building a classification model to detect health care fraud and abuse. The experimental results show that a great extent of redundant features can be eliminated by our feature selection algorithm, resulting in both accuracy improvement and computation cost reduction.

AB - Many techniques have recently been proposed for discovering structural patterns. Using the discovered structural patterns as features for classification has shown success in some application domains. However, the efficiency and effectiveness of such a classification algorithm is often impeded by the huge number of structural patterns discovered by the associated structural pattern mining algorithm. In this paper, we focus on the feature selection problem of structural patterns. The goal is to develop a scheme that effectively selects a subset of structural patterns as the features for the following induction algorithm. We show how to make use of the downward closure property inherent in the structural patterns to design a novel feature selection algorithm. We also evaluate our algorithm by applying the real-world health insurance data for building a classification model to detect health care fraud and abuse. The experimental results show that a great extent of redundant features can be eliminated by our feature selection algorithm, resulting in both accuracy improvement and computation cost reduction.

UR - http://www.scopus.com/inward/record.url?scp=27544483300&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=27544483300&partnerID=8YFLogxK

M3 - Conference contribution

SP - 55

BT - Proceedings of the Annual Hawaii International Conference on System Sciences

A2 - Spraque, Jr., R.H.

ER -