基于主成分分析和Logistic回归的上市公司财务困境预警模型的研究

摘要

Using the financial data of listed companies, the model of pre-warning in financial crisis is constructed combining PCA and Logistic regression method with help of the SPSS and MATLAB software in this paper. The results show that the model has good predictive effect. The model which is established based on the training samples is tested and the prediction results are basically consistent with the actual data. The errors are acceptable. This shows that the results of the financial pre-warning are entirely consistent with the actual financial status. That means the model has very good forecast accuracy and practical promotion.%采用主成分分析与Logistic回归相结合,利用上市公司财务数据,借助SPSS和MATLAB软件,构建了上市公司财务困境预警模型.研究结果表明:该模型具有很好的预测效果,根据训练样本建立的模型,经过检验,其预测结果基本与实际相符,误差水平可以接受.这说明财务预警结果几乎完全符合其所处的财务状态,同时也说明了该模型具有很好的预测准确度和实用推广性.