Intelligent Financial Warning Method Based on Support Vector Machines and Artificial Neural Networks
Classical financial distress forecasting methods are only focused on a certain trade of corporations of in our country, but not to deal with a concrete corporation financial condition. An intelligent financial distress warning model is proposed using Support Vector Machines (SVMs) and Artificial Neural Networks (ANNs) in this paper for a certain corporation. Key factors that influence corporations are selected using Factors Analysis Approach (FAA) for past years. Values of key factors of future years are estimated using SVMs for its well generalization ability. Known values of key factors and financial condition of past years are as samples to train ANNs. After training, using estimated key factors of future years by SVMs as inputs, the financial condition of future for a certain corporation can be predicted using ANNs. Using STzhujiang and Non-STshenzhenye stocks as samples, the objective is to make SVMs and ANNs as financial affairs early warning research tools by building an intelligent and individual financial distress warning model. The model built for individual industries would be even more predictive than general models built with multi-industry samples. Results show that SVMs and ANNs are valuable tools to make up of financial affairs early warning system and to forecast ST and Non-ST companies accurately whether they are being in financial distress. The companies can build their own financial distress forecasting patterns based on their own running surroundings using proposed financial affairs early warning models.
Financial Warning System Financial Ratios Factors Analysis Approach Support Vector Machines Artificial
Bo TIAN Zheng QIN
School of Management,Xian Jiaotong University,Xian 710049,China
国际会议
北京
英文
2007-05-30（万方平台首次上网日期，不代表论文的发表时间）