Developing expert system on decision making unit efficiency

Efficiency is a key concept for financial institutions. As personnel specifications have greatest impact on efficiency, they can help us designing work environments for maximizing efficiency. Providing information on multiple input and output factors are a complicated and time consuming procedure. Developing expert system in this situation is hard. This paper proposed a procedure that solved mentioned problem. At first, the integrated approach determining important attributes and then expert system is developed. The integrated approach uses Data Envelopment Analysis (DEA) and Data Mining tools. DEA is used for DMUs efficiency evaluation. Artificial Neural Network (ANN) and Cross Validation Test Technique (CVTT) are used for precision testing and forecasting and finally DEA is again utilized for identification of attributes importance. ANN is used for determining important attributes and developing expert system. As well, K-means algorithm is used in developing expert system. A Procedure is proposed to developing expert system with mentioned tools and completed rule base. The constructed expert system helps managers to forecast DMUs efficiencies by selected attributes and grouping inferred efficiency. Also, they can assess new situation before happening and compare with present situation. The proposed integrated approach is applied to an actual banking system and its superiorities and advantages are discussed.