摘要
研究企业财务困境预测问题,影响财务困境的输入变量参数较多,存在输入维数和冗余信息,造成预测效率低。如何准确选择合理的输入变量参数是提高财务困境预测精度的关键。为了解决财务困境输入变量选择不合理导致预测精度低等难题,提出采用主成分-遗传-支持向量机(PCA-GA-SVM)的企业财务困境组合预测方法。先用主成分分析法(PCA)合理选取企业财务困境的输入变量,然后结合遗传算法(GA)利用训练集的数据对SVM最优参数寻优,得到企业财务困境预测模型,最后采用具体企业财务数据进行仿真。实验结果表明,PCA-GA-SVM的企业财务困境预测方法提高了财务困境的预测精度。
Study of financial distress prediction problem. There are many variables affecting the financial dilem- ma, and choosing reasonable input variables is the key to improve the financial distress prediction accuracy. In order to solve the problem that the input variables selection is not reasonable, causing low precision of financial difficulties prediction, the paper put forward a corporate financial distress prediction method based on principal component analy- sis, genetic algorithm and support vector machine（ PCA -GA -SVM）. First, by using the principal component a- nalysis（PCA）, the input variables of enterprise＇s financial predicament were reasonably selected, and then com- bined with genetic algorithm （GA）, the SVM parameters were optimized with a set of training data, to obtain the en- terprise financial distress prediction model. Finally, the specific financial data of the enterprise were adopted to car- ries out the simulation experiment. The experimental results show that, the PCA - GA - SVM prediction method of enterprise financial distress can improve the financial distress prediction accuracy.
出处
《计算机仿真》
CSCD
北大核心
2012年第6期355-358,共4页
Computer Simulation
基金
江苏建筑职业技术学院院级科研课题(JYA309-08)
