基于遗传算法和神经网络的上市公司财务困境预测  

Financial distress predicting model based on genetic algorithm and neural network

作　　者：西凤茹[1] 时文超[1] 

机构地区：[1]辽宁科技大学工商管理学院,辽宁鞍山114051

出　　处：《辽宁科技大学学报》2013年第2期166-171,共6页Journal of University of Science and Technology Liaoning

基　　金：辽宁省教育厅资助项目(W2012045)

摘　　要：针对上市公司财务困境预测问题,以中国制造业上市公司为研究对象,以财务状况异常而被特别处理作为上市公司陷入财务困境的标志,将股权结构和董、监事会状况指标加入到财务预警的指标体系中,应用遗传算法优化的BP神经网络算法,并对独立检验样本集进行预测,将预测结果同logistic方法、支持向量机方法和BP神经网络方法进行比较。结果表明,GA-BP神经网络方法在提前两年和三年预测中,总正确率分别达到91.25%和82.5%,优于其他方法,具有较大的应用价值。To solve financial distress prediction problem,Chinese manufacturing listed companies are regarded as the research objects,the special treatment（ST） as abnormal financial situation is regarded as symbol that listed companies are in financial distress,and index of ownership structure、characteristic of directorate and supervisory board is added into index system of financial prediction,BP neural network model optimized by genetic algorithm is constructed and the prediction is carried out for independent testing data sets.The results are compared with the research of logistic model、support vector machine model and BP neural network.The results show that the accuracy ratios of GA-BP neural network model are 91.25% and 82.5% for two-year and three-year prediction respectively,which has a great significant application value.

关 键 词：财务困境 危机预测 遗传算法 神经网络 因子分析 

分 类 号：F224.9[经济管理—国民经济]