摘要
对企业经营绩效的评价主要采用线性和非线性两类评价模型,非线性评价模型能够更好地对经济现象进行仿真,评价结果客观、准确,更加具有实际参考价值。本文从考核投入产出效率的角度出发,选取基本财务指标构成评价体系,在此基础上,建立基于误差逆传播人工神经网络(BP神经网络)的高新技术企业绩效评价模型。以医药行业2003年度22家上市公司财务数据作为神经网络的训练和测试样本,将训练好的BP神经网络应用于企业绩效的当期评价和仿真预测,实证分析结果令人满意。
Two categories of evaluation model, linear model and non-linear model have been applied to evaluating enterprise performance frequently. The simulation efficiency of economic phenomena through the application of non-linear rnodel is satisfactory. The result of evaluation is objective, scientific and more valuable for administrators as reference. From the view of evaluating input-output efficiency, the article constructs integrated indexes system by selecting basic financial indexes. The back propagation neural network method is applied to the integrated evaluation based on the 2003 annual financial data of twenty-two listed companies in medicine industry. It can evaluate the enterprise performance and forecast prospect through training formal fingers. The results show that the application of BP neutral networks to evaluating the enterprise performance is satisfactory.
出处
《运筹与管理》
CSCD
2006年第4期137-140,共4页
Operations Research and Management Science
基金
教育部博士点基金资助项目(20040217005)
国防科工委技术基础资助项目(C192005C001)
黑龙江省自然科学基金资助项目(2004-25)
关键词
运筹学
综合评价
BP神经网络
混合IDEA模型
经营绩效
operations research
integrated evaluation
BP neural networks
mixed DEA model
enterprise performance
