(A) neural network approach to management evaluation and feedback adjustment

Abstract

Up to the present, the evaluation of managerial operations and the feedback adjustment based on the evaluated results have been performed by managers and related experts with their domain knowledge. However, the conventional approach has some limitations : too much expenses in cost and time, and inconsistencies among cases. These drawbacks inherently come from the fact that the evaluation process is troublesome to be modeled, because the knowledge is difficult to be represented as a well-defined form. The recent advance in artificial neural networks makes it possible to efficiently represent the complicated and unstructured knowledge. In this study, the management evaluation model which utilizes artificial neural networks and the feedback adjustment process which uses sensitivity analysis to the evaluation model are proposed. Based on the model and the process, the Management Evaluation and Feedback Adjustment System (MEFAS) is develop, using C language on the IBM PC. MEFAS is demonstrated in the case study of Korea Mobile Telecommunications Corporation(KMTC) and validated by comparing experts`` evaluated results with MEFAS.