关键词: 财务危机预警, 财务指标, 自组织映射, 模糊C均值, 隶属度

Abstract: The good financial distress prediction model can effectively monitor the enterprise situation, which can avoid the tragedy for enterprises, such as business failures, restructuring and so on. The purpose in this paper is to construct the fuzzy self-organizing map model(FSOM), which integrates self-organizing map(SOM)with fuzzy C-Means(FCM), and apply it to financial distress prediction. This new model introduces the fuzzy membership to the learning rate function and automatically updates the neighborhood of the winner node in the process of the calculation. The batch algorithm is adopted in this paper in the process of iterative calculation. Therefore, FSOM model can enhance prediction accuracy and acquire the stable outcomes. In the empirical study, we research financial indicators of listed companies which are from the Shanghai Stock Exchange and the Shenzhen Stock Exchange in China. FSOM model has a better prediction performance in terms of financial distress prediction by comparing other traditional prediction models.

Key words: financial distress prediction, financial indicator, self-organizing map, fuzzy c-means, fuzzy membership