我国部分上市公司两种财务危机预警模型的比较研究  

A COMPARATIVE STUDY ON TWO FINANCIAL CRISIS WARNING MODELS OF OUR COUNTRY'S PARTIAL LISTED COMPANIES

作　　者：上官鸣[1] 马邵利[1] 陈小超[1] 

机构地区：[1]陕西科技大学管理学院,陕西西安710021

出　　处：《陕西科技大学学报（自然科学版）》2008年第2期143-146,150,共5页Journal of Shaanxi University of Science & Technology

摘　　要：财务危机预警模型是防范财务危机的有效途径之一.采用主成分分析模型和BP算法神经网络仿真模型对我国部分上市公司进行了财务困境预测,以我国上市公司中60家公司的财务指标作为估计样本组,以40家公司作为预测样本组建立了财务危机预警模型.对两种模型的计算分析表明,BP算法神经网络仿真模型的预测准确率明显优于主成分分析模型,具有广泛的适用性.Financial crises warning model is an effective way to prevent financial crises. It uses the principal component analysis model andthe BP neural network model to forecast financial difficult position in our listed companies. It uses the cross section financial indexes of 60 listed companies as estimate samples and the corresponding 40 listed companies as testing samples. The result indicates that the BP neural network model is better than the principal components analysis model on the forecasting rate accurately,which will be used in large areas.

关 键 词：主成分分析 BP算法神经网络仿真模型 财务指标 模型 

分 类 号：F234.4[经济管理—会计学] O212.1[经济管理—国民经济]