A Case-Based Reasoning System for Indirect Bank Lending

Authors:
Atish P. Sinha;Mark A. Richardson
Affiliations:
University of Dayton, Dayton, OH 45469-2130, USA;522 Eagle Street, Petaluma, CA 94954, USA
Venue:
International Journal of Intelligent Systems in Accounting and Finance Management
Year:
1996

Citing 0
Cited 1

Evaluating and Tuning Predictive Data Mining Models Using Receiver Operating Characteristic Curves

Journal of Management Information Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

By implementing case-based reasoning (CBR) systems, business organizations can utilize past cases—a key data resource—for future decision making. CBR is particularly suitable for business domains that have available a large amount of historical data. One such domain is indirect bank lending. In this paper, we present a case-based system that operates in the bank lending domain. The system recommends whether an indirect loan application should be approved or denied, based on past experiences. We describe how the system was developed and explain how the system functions. The system was empirically evaluated using actual loan cases. The positive results of the evaluation confirm our hypothesis that CBR is an attractive decision-making methodology for the bank lending domain. © 1996 Wiley Periodicals, Inc.