Application of classification methods to individual disability income insurance fraud detection

Yi Peng, Gang Kou, Alan Sabatka, Jeff Matza, Zhengxin Chen, Deepak Khazanchi, Yong Shi

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

9 Citations (Scopus)

Abstract

As the number of electronic insurance claims increases each year, it is difficult to detect insurance fraud in a timely manner by manual methods alone. The objective of this study is to use classification modeling techniques to identify suspicious policies to assist manual inspections. The predictive models can label high-risk policies and help investigators to focus on suspicious records and accelerate the claim-handling process. The study uses health insurance data with some known suspicious and normal policies. These known policies are used to train the predictive models. Missing values and irrelevant variables are removed before building predictive models. Three predictive models: Naïve Bayes (NB), decision tree, and Multiple Criteria Linear Programming (MCLP), are trained using the claim data. Experimental study shows that NB outperformed decision tree and MCLP in terms of classification accuracy.

Original languageEnglish (US)
Title of host publicationComputational Science - ICCS 2007 - 7th International Conference, Proceedings
Pages852-858
Number of pages7
EditionPART 3
StatePublished - Dec 1 2007
Event7th International Conference on Computational Science, ICCS 2007 - Beijing, China
Duration: May 27 2007May 30 2007

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 3
Volume4489 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference7th International Conference on Computational Science, ICCS 2007
CountryChina
CityBeijing
Period5/27/075/30/07

Fingerprint

Disability Insurance
Fraud Detection
Fraud
Disability
Insurance
Predictive Model
Linear Programming
Decision Trees
Multiple Criteria
Bayes
Decision trees
Decision tree
Linear programming
Health insurance
Missing Values
Health Insurance
Accelerate
Inspection
Labels
Experimental Study

Keywords

  • Classification
  • Decision tree
  • Insurance fraud detection
  • Multiple Criteria Linear Programming (MCLP)
  • Naïve Bayes (NB)

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Peng, Y., Kou, G., Sabatka, A., Matza, J., Chen, Z., Khazanchi, D., & Shi, Y. (2007). Application of classification methods to individual disability income insurance fraud detection. In Computational Science - ICCS 2007 - 7th International Conference, Proceedings (PART 3 ed., pp. 852-858). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4489 LNCS, No. PART 3).

Application of classification methods to individual disability income insurance fraud detection. / Peng, Yi; Kou, Gang; Sabatka, Alan; Matza, Jeff; Chen, Zhengxin; Khazanchi, Deepak; Shi, Yong.

Computational Science - ICCS 2007 - 7th International Conference, Proceedings. PART 3. ed. 2007. p. 852-858 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4489 LNCS, No. PART 3).

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

Peng, Y, Kou, G, Sabatka, A, Matza, J, Chen, Z, Khazanchi, D & Shi, Y 2007, Application of classification methods to individual disability income insurance fraud detection. in Computational Science - ICCS 2007 - 7th International Conference, Proceedings. PART 3 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 3, vol. 4489 LNCS, pp. 852-858, 7th International Conference on Computational Science, ICCS 2007, Beijing, China, 5/27/07.
Peng Y, Kou G, Sabatka A, Matza J, Chen Z, Khazanchi D et al. Application of classification methods to individual disability income insurance fraud detection. In Computational Science - ICCS 2007 - 7th International Conference, Proceedings. PART 3 ed. 2007. p. 852-858. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3).
Peng, Yi ; Kou, Gang ; Sabatka, Alan ; Matza, Jeff ; Chen, Zhengxin ; Khazanchi, Deepak ; Shi, Yong. / Application of classification methods to individual disability income insurance fraud detection. Computational Science - ICCS 2007 - 7th International Conference, Proceedings. PART 3. ed. 2007. pp. 852-858 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3).
@inproceedings{d2a7ecb2c4dd421f96860647a41943bb,
title = "Application of classification methods to individual disability income insurance fraud detection",
abstract = "As the number of electronic insurance claims increases each year, it is difficult to detect insurance fraud in a timely manner by manual methods alone. The objective of this study is to use classification modeling techniques to identify suspicious policies to assist manual inspections. The predictive models can label high-risk policies and help investigators to focus on suspicious records and accelerate the claim-handling process. The study uses health insurance data with some known suspicious and normal policies. These known policies are used to train the predictive models. Missing values and irrelevant variables are removed before building predictive models. Three predictive models: Na{\"i}ve Bayes (NB), decision tree, and Multiple Criteria Linear Programming (MCLP), are trained using the claim data. Experimental study shows that NB outperformed decision tree and MCLP in terms of classification accuracy.",
keywords = "Classification, Decision tree, Insurance fraud detection, Multiple Criteria Linear Programming (MCLP), Na{\"i}ve Bayes (NB)",
author = "Yi Peng and Gang Kou and Alan Sabatka and Jeff Matza and Zhengxin Chen and Deepak Khazanchi and Yong Shi",
year = "2007",
month = "12",
day = "1",
language = "English (US)",
isbn = "9783540725879",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 3",
pages = "852--858",
booktitle = "Computational Science - ICCS 2007 - 7th International Conference, Proceedings",
edition = "PART 3",

}

TY - GEN

T1 - Application of classification methods to individual disability income insurance fraud detection

AU - Peng, Yi

AU - Kou, Gang

AU - Sabatka, Alan

AU - Matza, Jeff

AU - Chen, Zhengxin

AU - Khazanchi, Deepak

AU - Shi, Yong

PY - 2007/12/1

Y1 - 2007/12/1

N2 - As the number of electronic insurance claims increases each year, it is difficult to detect insurance fraud in a timely manner by manual methods alone. The objective of this study is to use classification modeling techniques to identify suspicious policies to assist manual inspections. The predictive models can label high-risk policies and help investigators to focus on suspicious records and accelerate the claim-handling process. The study uses health insurance data with some known suspicious and normal policies. These known policies are used to train the predictive models. Missing values and irrelevant variables are removed before building predictive models. Three predictive models: Naïve Bayes (NB), decision tree, and Multiple Criteria Linear Programming (MCLP), are trained using the claim data. Experimental study shows that NB outperformed decision tree and MCLP in terms of classification accuracy.

AB - As the number of electronic insurance claims increases each year, it is difficult to detect insurance fraud in a timely manner by manual methods alone. The objective of this study is to use classification modeling techniques to identify suspicious policies to assist manual inspections. The predictive models can label high-risk policies and help investigators to focus on suspicious records and accelerate the claim-handling process. The study uses health insurance data with some known suspicious and normal policies. These known policies are used to train the predictive models. Missing values and irrelevant variables are removed before building predictive models. Three predictive models: Naïve Bayes (NB), decision tree, and Multiple Criteria Linear Programming (MCLP), are trained using the claim data. Experimental study shows that NB outperformed decision tree and MCLP in terms of classification accuracy.

KW - Classification

KW - Decision tree

KW - Insurance fraud detection

KW - Multiple Criteria Linear Programming (MCLP)

KW - Naïve Bayes (NB)

UR - http://www.scopus.com/inward/record.url?scp=38149026389&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38149026389&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:38149026389

SN - 9783540725879

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 852

EP - 858

BT - Computational Science - ICCS 2007 - 7th International Conference, Proceedings

ER -