电力工业的非参数随机前沿面分析

随机前沿面模型是评价决策单元投入产出效率的重要方法之一,但模型中参数的求解和误差项的分解比较困难,寻求简洁有效算法是研究的重点。对随机前沿面模型构造过程中误差项的处理进行了研究,采用非参数方法对随机前沿面模型的误差项进行处理,然后利用最小二乘法进行回归分析并建立随机前沿面模型,有效地解决了随机前沿面模型中误差项分解和计算的困难。电力工业的实证研究表明,基于非参数处理的随机前沿面模型给出了具体函数形式,模型求解变得直观和容易,而且效率评价值靠前的企业是那些成功建立现代企业制度的企业,说明了方法的合理性。

Nonparametric method-stochastic frontier analysis of electric power industry

Stochastic frontier analysis is an effective method to analyze the efficiency of Decision Making Unit(DMU),but the decomposition of the error term is difficult,the research aims at searching for concise and efficient algorithm.Adopting nonparametric method to dispose the error term of stochastic frontier model,regressiona1 analysis of stochastic frontier model was constructed using Least Square Method.The method can resolve the difficulty of the decomposition of the error.Empirical research of electric power industry shows that the nonparametric method-stochastic frontier analysis can give the function form and make solving of parameters become easier,and electric power companies with high technical efficiency are those which had developed modern enterprise system successfully,which proves the rationality of the method.