Cluster analysis in retail segmentation for credit scoring

Authors

  • Sanja Scitovski J.J. Strossmayer University of Osijek
  • Nataša Šarlija Faculty of Economics J.J. Strossmayer University of Osijek

Abstract

The aim of this paper is to segment retail clients by using adaptive Mahalanobis clustering in a way that each segment can be suitable for separate credit scoring development such that a better risk assessment of retail clients could be accomplished. A real data set on retail clients from a Croatian bank was used in the paper. Grouping of the data point set is carried out by using the adaptive Mahalanobis partitioning algorithm (see, e.g., [20]). It is an incremental algorithm, which recognizes ellipsoidal clusters with the main axes in the directions of eigenvectors of the corresponding covariance matrix of the data set. On the basis of the given data set, by using the well-known DIRECT algorithm for global optimization it is possible to search successively for an optimal partition with
k=2,3,... clusters. After that, a partition with the most appropriate number of clusters is determined by using various validity indexes. Based on the description of each cluster,
banks could decide to develop a separate credit scoring model for each cluster as well as to create a business strategy customized to each cluster.

Downloads

Additional Files

Published

2015-01-17

Issue

Volume 5 (2014), Number 2

Section

CRORR Journal Regular Issue

License

- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal

- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.

- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).