Font Size: a A A

Study On The Construction And Application Of Production Frontier

Posted on:2004-01-20Degree:DoctorType:Dissertation
Country:ChinaCandidate:J X WangFull Text:PDF
GTID:1116360092480649Subject:Management decision-making and logistics technology
Abstract/Summary:PDF Full Text Request
The enterprise, which aim is to making profit, is in pursuit of profit maximization and optimal production situation. Production possible and optimal frontier described by theoretical function is called production frontier. On the ground of domestic and overseas existing research, from the points of parametric estimation method, non-parametric estimation method and the combination of these two methods, the theory of production frontier is studied comprehensively and elaborately. The main content of the dissertation is as following:1.For the difficulty of the error decomposing of the stochastic frontier model, maximum likelihood estimation is adopted to construct the solving model, and the error decomposing is solved satisfactorily in the empirical research. In addition, other distribution assumptions of management error of stochastic frontier model are discussed intensively, and distribution assumption is simulated, the selecting value principle of distribution assumption is analyzed from the simulating results.2.The characteristics of Aigner-chu algorithm and probabilistic frontier are analyzed, the calculating process of Aigner-chu algorithm is illustrated with an example. In addition, adopting genetic algorithm to search the parameters of parametric estimation method is put forward. For the function form assumption of parametric estimation method is subjective, adopting different function form to construct production frontier is analyzed, the concepts such as Tle(total loss of efficiency ) and Ale(average loss of efficiency) etc. are put forward and the empirical research is done.3.On the base of non-parametric estimation method, the efficiency measurement is researched. With an example of electric power industry, pure technique-efficiency, size-efficiency, input disposition-degree, input component-efficiency and the input combination- efficiency are analyzed in this dissertation, and the calculating results are optimized and analyzed in the end.4.The good qualities of parametric estimation method and non-parametric estimation method are extracted to construct production frontier. Adopting non-parametric estimation method to handle the management error of stochastic frontier is discussed first, and the difference between this method and non-parametric estimation method is analyzed. For the negative and invalid output data exists universally in reality, the method which adopts non-parametric estimation method to filtrate the data is put forward and illustrated with an example, the characteristics of the method is analyzed at last.
Keywords/Search Tags:parametric estimation method, non-parametric estimation method, efficiency, genetic algorithm, electric power industry, Monte-carlo, simulation
PDF Full Text Request
Related items