“…Figure 1 shows an example of FNN composed of n input nodes, m nodes in one hidden layer, and k output nodes. FNN is widely employed as a classification tool in a variety of application domains, including medical analysis, credit scoring, pattern recognition, speech recognition, handwriting recognition, product inspection, drug discovery and development, biological classification, natural language processing, document classification, and network security [6,8,20]. In credit risk assessment, FNN is used to develop credit risk models from the historical data and predict the future corporate bankruptcies [1,2,27].…”