Improving scoring techniques in banking applications

Pascal Perrodo

Applied Stochastic Models and Data Analysis, 1993, vol. 9, issue 3, 231-244

Abstract: Scoring techniques are largely used today in banking applications, for marketing as well as risk assessment or control purposes. The most widely used method is Fisher's discriminatory analysis. We discuss here this method from a technical standpoint, showing its weaknesses and then propose a simpler and more general alternative. We first make an original presentation of the various scoring techniques and their basic principles. In particular, we provide a detailed description of Fisher's method, insisting on sensitive steps such as the discretization of continuous variables which is responsible for the nonlinearity of the model. We then go on to propose a new method that is completely linear and improves scoring performance. This technique is easy to explain, compute and use in real applications. Finally, we compare the performance of this new method with that of traditional techniques and of the more up‐to‐date neural nets method.

Date: 1993
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://doi.org/10.1002/asm.3150090304

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:wly:apsmda:v:9:y:1993:i:3:p:231-244

Access Statistics for this article

More articles in Applied Stochastic Models and Data Analysis from John Wiley & Sons
Bibliographic data for series maintained by Wiley Content Delivery ().