基于邻域粗糙集属性约简的对偶约束式LS-SVM财务困境预测模型研究  

Study on Financial Distress Prediction Model of Least Squares Support Vector Machine of Dual Constraint Type Based on Attribute Reduction of Neighborhood Rough Set

作　　者：赵冠华[1] 

机构地区：[1]山东财政学院会计学院,山东济南250014

出　　处：《运筹与管理》2011年第3期132-139,共8页Operations Research and Management Science

基　　金：国家自然科学基金资助项目(70840018);山东省科技攻关计划项目(2008GG30009005);山东省软科学研究计划项目(2008RKA223)

摘　　要：为了提高财务困境预测的正确率,改善模型预测的效果,将邻域粗糙集和遗传算法应用于对偶约束式最小二乘支持向量机,提出了一种基于邻域粗糙集属性约简的对偶约束式最小二乘支持向量机预测模型。同时,给出了这一改进模型的实现步骤。实证结果表明,通过邻域粗糙集指标预处理和遗传算法参数优化后,不但提高了模型预测的正确率,还降低了模型运行的时间,证实了该模型应用于财务困境预测是有效的。In order to increase the accuracy of financial distress prediction and improve the prediction effect of model,this paper applies neighborhood rough set and genetic algorithm to least squares support vector machine of dual constraint type and advances a prediction model of least squares support vector machine of dual constraint type which is based on attribute reduction of neighborhood rough set.Besides,it presents the procedures of carrying out the improved model.The experimental results show that the model increases its prediction accuracy and reduce its running time by pretreating indicators with neighborhood rough set and optimizating parameters with genetic algorithm.The model is effective in forecasting financial distress.

关 键 词：邻域粗糙集 对偶约束 最小二乘支持向量机 遗传算法 财务困境预测 

分 类 号：F253.7[经济管理—国民经济]