Studying Impact of Decision Making Units Features on Efficiency by Integration of Data Envelopment Analysis and Data Mining Tools

Abstract

Generally, data mining is the process of analyzing data from different viewpoint and summarizing it into valuable information. This area presents new theories and methods for processing large volumes of data and has obtained noteworthy consideration among researchers. In this paper, a new approach for decision-making process is developed based on the rough set theory of data mining and neural networks combined with data envelopment analysis method. The proposed procedure assesses the effect of personnel attributes on efficiency, utilizing DEA tool in estimating the efficiency of alternative decision making unites. By developing decision system, rough set theory is applied for feature selection (reducts) and all of plausible and meaningful ANN models are constructed for each reduct. Finally DEA method is used for selecting the best reduct and also most important personnel attributes for efficiency analysis. Persian bank branches employed for data generation and the characteristics of its personnel are analyzed on effectiveness of bank branches.