The evaluation of a firm's performance has been extensively studied in the field of accounting. This paper presents an integrated devaluation system which is based on a neural-network approach designed to predict a firm's performance. The integrated system combines two neural network paradigms – self-organizing network and Grossberg's instar network.
The results of our experimental studies reveal that the proposed integrated system exhibits reliable classifying potential with a high degree of fault tolerance in the overall process of evaluating a firm's performance.
Journal of Computer Information Systems 36(3), pp.10-15