Application of the Differential Evolution Algorithm to Group a Bank's Individual Clients
Application of the Differential Evolution Algorithm to Group a Bank's Individual Clients

Summary/Abstract: Objective: The aim of the article is to present the results of grouping individual clients of a bank with the differential evolution algorithm. Research Design & Methods: The research offers conclusions based on analysis of the bank’s customer base and deductive and inductive reasoning. Findings: The results of the authors’ research show that the differential evolution algorithm correctly groups bank customers and can be used for this purpose. Implications/Recommendations: The differential evolution algorithm is an alternative to the commonly used k-means algorithm. The algorithm generates several competing solutions in one iteration. It enables independence from starting vectors and greater effectiveness in searching for an optimal solution. The differential evolution algorithm was itself enriched with a variable that allows the optimal number of clusters to be selected. Each iteration contained proposed solutions (chromosomes) that were evaluated by the target function built on the CS measure proposed by Chou. Contribution: The article presents the application of the differential evolution algorithm to group a bank’s clients.

• Details
• Contents