What is it about?

We propose a model that utilizes for each numerical feature of loan granting case various similarity measures. These similarity measures are in the form of preference functions which allow the integration of expert knowledge in the process of similarity measurement. This way the famous kNN algorithm is extended to work with integrated domain knowledge.

Featured Image

Why is it important?

It allows the integration of expert knowledge in the process of similarity measurement, which improves the standard way of similarity measurement, which measures all features with the same similarity measure.

Perspectives

This paper gave a proposal of a model that was later extended and published in Expert Systems and Applications and journals by my PhD student Sanja Vukovic "A case-based reasoning model that uses preference theory functions for credit scoring" The paper could be extended in various directions, one being to enable the model to work with categorical and mixed data types as well.

Boris Delibasic
University of Belgrade - Faculty of Organizational Sciences

Read the Original

This page is a summary of: Loan Granting Knowledge System, Journal of Decision System, January 2006, Taylor & Francis,
DOI: 10.3166/jds.15.309-329.
You can read the full text:

Read

Contributors

The following have contributed to this page