Machine learning Clustering Algorithms Based on the DEA Optimization Approach for Banking System in Developing Countries

Abstract

Machine learning grows quickly, which has made numerous academic discoveries and is extensively evaluated in several areas. Optimization, as a vital part of machine learning, has fascinated much consideration of practitioners. The primary purpose of this paper is to combine optimization and machine learning to extract hidden rules, remove unrelated data, introduce the most productive Decision-Making Units (DMUs) in the optimization part, and to introduce the algorithm with the highest accuracy in Machine learning part. In the optimization part, we evaluate the productivity of 30 banks from eight developing countries over the period 2015-2019 by utilizing Data Envelopment Analysis (DEA). An additive Data Envelopment Analysis (DEA) model for measuring the efficiency of decision processes is used. The additive models are often named Slack Based Measure (SBM). This group of models measures efficiency via slack variables. After applying the proposed model, the Malmquist Productivity Index (MPI) is computed to evaluate the productivity of companies. In the machine learning part, we use a specific two-layer data mining filtering pre-processes for clustering algorithms to increase the efficiency and to find the superior algorithm. This study tackles data and methodology-related issues in measuring the productivity of the banks in developing countries and highlights the significance of DMUs productivity and algorithms accuracy in the banking industry by comparing suggested models.