A multiple intelligent agent system for credit risk prediction via an optimization of localized generalization error with diversity

S., Daniel YEUNG, W. Y., Wing NG, P. F., Aki CHAN, P. K., Patrick CHAN, Michael Arthur FIRTH, C. C., Eric TSANG

Research output: Journal PublicationsJournal Article (refereed)Researchpeer-review

20 Citations (Scopus)

Abstract

Company bankruptcies cost billions of dollars in losses to banks each year. Thus credit risk prediction is a critical part of a bank's loan approval decision process. Traditional financial models for credit risk prediction are no longer adequate for describing today's complex relationship between the financial health and potential bankruptcy of a company. In this work, a multiple classifier system (embedded in a multiple intelligent agent system) is proposed to predict the financial health of a company. In our model, each individual agent (classifier) makes a prediction on the likelihood of credit risk based on only partial information of the company. Each of the agents is an expert, but has limited knowledge (represented by features) about the company. The decisions of all agents are combined together to form a final credit risk prediction. Experiments show that our model out-performs other existing methods using the benchmarking Compustat American Corporations dataset.
Original languageEnglish
Pages (from-to)166-180
Number of pages15
JournalJournal of Systems Science and Systems Engineering
Volume16
Issue number2
DOIs
Publication statusPublished - 1 Jan 2007
Externally publishedYes

Fingerprint

Intelligent agents
Industry
Classifiers
Health
Benchmarking
Embedded systems
Costs
Experiments

Keywords

  • Business intelligence
  • Credit rating
  • Feature grouping
  • Localized generalization error
  • Multiple classifier system

Cite this

YEUNG, S., Daniel ; NG, W. Y., Wing ; CHAN, P. F., Aki ; CHAN, P. K., Patrick ; FIRTH, Michael Arthur ; TSANG, C. C., Eric. / A multiple intelligent agent system for credit risk prediction via an optimization of localized generalization error with diversity. In: Journal of Systems Science and Systems Engineering. 2007 ; Vol. 16, No. 2. pp. 166-180.
@article{24a175a519764731af4fbb67bc49ee33,
title = "A multiple intelligent agent system for credit risk prediction via an optimization of localized generalization error with diversity",
abstract = "Company bankruptcies cost billions of dollars in losses to banks each year. Thus credit risk prediction is a critical part of a bank's loan approval decision process. Traditional financial models for credit risk prediction are no longer adequate for describing today's complex relationship between the financial health and potential bankruptcy of a company. In this work, a multiple classifier system (embedded in a multiple intelligent agent system) is proposed to predict the financial health of a company. In our model, each individual agent (classifier) makes a prediction on the likelihood of credit risk based on only partial information of the company. Each of the agents is an expert, but has limited knowledge (represented by features) about the company. The decisions of all agents are combined together to form a final credit risk prediction. Experiments show that our model out-performs other existing methods using the benchmarking Compustat American Corporations dataset.",
keywords = "Business intelligence, Credit rating, Feature grouping, Localized generalization error, Multiple classifier system",
author = "YEUNG, {S., Daniel} and NG, {W. Y., Wing} and CHAN, {P. F., Aki} and CHAN, {P. K., Patrick} and FIRTH, {Michael Arthur} and TSANG, {C. C., Eric}",
year = "2007",
month = "1",
day = "1",
doi = "10.1007/s11518-007-5048-4",
language = "English",
volume = "16",
pages = "166--180",
journal = "Journal of Systems Science and Systems Engineering",
issn = "1004-3756",
publisher = "Springer-Verlag GmbH and Co. KG",
number = "2",

}

A multiple intelligent agent system for credit risk prediction via an optimization of localized generalization error with diversity. / YEUNG, S., Daniel; NG, W. Y., Wing; CHAN, P. F., Aki; CHAN, P. K., Patrick; FIRTH, Michael Arthur; TSANG, C. C., Eric.

In: Journal of Systems Science and Systems Engineering, Vol. 16, No. 2, 01.01.2007, p. 166-180.

Research output: Journal PublicationsJournal Article (refereed)Researchpeer-review

TY - JOUR

T1 - A multiple intelligent agent system for credit risk prediction via an optimization of localized generalization error with diversity

AU - YEUNG, S., Daniel

AU - NG, W. Y., Wing

AU - CHAN, P. F., Aki

AU - CHAN, P. K., Patrick

AU - FIRTH, Michael Arthur

AU - TSANG, C. C., Eric

PY - 2007/1/1

Y1 - 2007/1/1

N2 - Company bankruptcies cost billions of dollars in losses to banks each year. Thus credit risk prediction is a critical part of a bank's loan approval decision process. Traditional financial models for credit risk prediction are no longer adequate for describing today's complex relationship between the financial health and potential bankruptcy of a company. In this work, a multiple classifier system (embedded in a multiple intelligent agent system) is proposed to predict the financial health of a company. In our model, each individual agent (classifier) makes a prediction on the likelihood of credit risk based on only partial information of the company. Each of the agents is an expert, but has limited knowledge (represented by features) about the company. The decisions of all agents are combined together to form a final credit risk prediction. Experiments show that our model out-performs other existing methods using the benchmarking Compustat American Corporations dataset.

AB - Company bankruptcies cost billions of dollars in losses to banks each year. Thus credit risk prediction is a critical part of a bank's loan approval decision process. Traditional financial models for credit risk prediction are no longer adequate for describing today's complex relationship between the financial health and potential bankruptcy of a company. In this work, a multiple classifier system (embedded in a multiple intelligent agent system) is proposed to predict the financial health of a company. In our model, each individual agent (classifier) makes a prediction on the likelihood of credit risk based on only partial information of the company. Each of the agents is an expert, but has limited knowledge (represented by features) about the company. The decisions of all agents are combined together to form a final credit risk prediction. Experiments show that our model out-performs other existing methods using the benchmarking Compustat American Corporations dataset.

KW - Business intelligence

KW - Credit rating

KW - Feature grouping

KW - Localized generalization error

KW - Multiple classifier system

UR - http://commons.ln.edu.hk/sw_master/2264

U2 - 10.1007/s11518-007-5048-4

DO - 10.1007/s11518-007-5048-4

M3 - Journal Article (refereed)

VL - 16

SP - 166

EP - 180

JO - Journal of Systems Science and Systems Engineering

JF - Journal of Systems Science and Systems Engineering

SN - 1004-3756

IS - 2

ER -