摘要
该文从财务困境概念漂移的全新视角,提出了基于滚动时间窗口支持向量机(support vector machine,SVM)的财务困境预测动态建模新方法。设计了面向概念漂移进行财务困境预测动态建模的思路框架,分为宽度固定的滚动时间窗口SVM和宽度可变的滚动时间窗口SVM分别展开算法设计。以中国上市公司为对象,通过模拟时间推移过程,对2000至2008期间被ST的上市公司及其配对公司共692个样本展开实证研究。结果表明:基于滚动时间窗口SVM的财务困境预测动态建模方法能够有效地适应财务困境的概念漂移现象,对未来企业财务困境的预测效果明显优于静态SVM模型。通过比较分析,认为适应性可变时间窗口SVM动态建模方法具有较好的应用推广性。
From the new view of financial distress concept drift,this proposes a new method for dynamic financial distress prediction modeling based on rolling time window support vector machine（SVM）.The framework of dynamic financial distress prediction modeling for handing concept drift is designed.The algorithms are designed respectively for two types of rolling time window SVM： fixed window size and alterable window size.With totally 692 samples from Chinese listed companies,which include ST companies from 2000 to 2008 and their paired non-ST companies,the empirical study is carried out by simulating the process of time passage.The results indicate that the proposed dynamic modeling methods based on rolling time window SVM can effectively adapt to the concept drift of financial distress.They significantly outperform the static SVM models in predicting future financial distress.By comparison,the rolling time window SVM with adaptable window size has better practicability.
出处
《管理工程学报》
CSSCI
北大核心
2010年第4期174-180,92,共8页
Journal of Industrial Engineering and Engineering Management
基金
国家自然科学基金项目(70801054)
浙江省自然科学基金项目(Y6090392)
关键词
财务困境预测
概念漂移
滚动时间窗口
支持向量机
financial distress prediction
concept drift
rolling time window
support vector machine
