A self tuning model for risk estimation

被引：11

作者：

Elliott, Robert J. ^{[1
]}

Filinkov, Alexei ^{[2
]}

机构：

[1] Univ Calgary, Haskayne Sch Business, Calgary, AB T2N 1N4, Canada

[2] Univ Adelaide, Sch Math Sci, Adelaide, SA 5005, Australia

来源：

EXPERT SYSTEMS WITH APPLICATIONS | 2008年 / 34卷 / 03期

关键词：

risk; estimation; hidden markov models;

D O I：

10.1016/j.eswa.2007.01.044

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Credit scoring models often use linear or logistic regression to investigate the relation between observed characteristics and credit ratings. The basic relation is, however, a form of Bayes' theorem. This paper proposes a model in which estimation techniques from hidden Markov models are adapted to evaluate the parameters of a risk profile. The risk being estimated might be financial, as in credit scoring, or alternatively whether an observed member of a population might represent some terrorist threat. (C) 2007 Elsevier Ltd. All rights reserved.

引用

页码：1692 / 1697

页数：6

共 7 条

[1]

BAESENS B, 2003, THESIS KATHOLIEKE U

[2]

Elliott R., 1995, Hidden Markov Models-Estimation and Control, V29

[3] EXACT ADAPTIVE FILTERS FOR MARKOV-CHAINS OBSERVED IN GAUSSIAN-NOISE [J].

ELLIOTT, RJ .

AUTOMATICA, 1994, 30 (09) :1399-1408

[4]

Lewis E., 1992, INTRO CREDIT SCORING

[5]

Ringrose T. J., 1997, Construction and Assessment of Classification Rules, V53, DOI https://doi.org/10.2307/2533581

[6] Modeling consumer acceptance probabilities [J].

Thomas, LC ;

Jung, KM ;

Thomas, SD ;

Wu, Y .

EXPERT SYSTEMS WITH APPLICATIONS, 2006, 30 (03) :499-506

[7]

THOMAS LC, 2002, CREDIT SCORING APPL

← 1 →