Study of Banking Customers Credit Scoring Indicators Using Artificial Intelligence and Delphi Method

Abstract

With the importance of lending in the banking industry, it is very important to use the indicators affecting credit to decide on lending. The purpose of the present study is to identify and prioritize the effective features in customer accreditation using the viewpoints of bank experts in Kerman and to compare them with existing indicators in models extracted from Meta-Heuristic and Artificial Intelligence methods. The aim is to find out whether there is a match between the human views that arise from knowledge and experience and the views of artificial intelligence that look at the problem as black-box modeling. Required data were collected by questionnaire method and Quantum Binary particle swarm optimization algorithm and analyzed by Delphi. The results show that the selected indices have 80% overlap between the two methods. Due to the results of research and high accuracy of artificial intelligence techniques, it is suggested that in order to give credit to customers in banks and financial and credit institutions, to consider a higher weight for these indicators.

Keywords

• keywords: credit scoring
• delphi
• meta-heuristic algorithm
• pattern recognition
• feature selection