Sensitivity Analysis On Fuzzy DEA-DA Model And DEA-Benchmarking Model

Data Envelopment Analysis is a method which based on the concept of relative efficiency of a multi-objective decision-making. Currently, the research of DEA is more mature in the certain environment, but the relatively application of DEA is still lacking in the fuzzy environment, the sensitivity analysis of FDEA-DA model are fewer in number. The author of this paper participates in the project of Studies on Enterprise Performance evaluation of DEA-Benchmarking research methods and dynamic game, which is authorized by the country natural science commission and Studies on Fuzzy Pattern Recognition Algorithm and its application based on DEA-DA, which is the Scientific and technological development projects in Jilin Province and consults a lot of academic documents about DEA Model, DEA-DA Model, Benchmarking and these Sensitivity Analysis in China and abroad. The theoretical knowledge of mathematical programming, fuzzy math, and mathematical statistics is comprehensively used; the sensitivity analysis on the FDEA-DA model and DEA-Benchmarking model is depth research and analysis in the paper. At last, the conclusions of the model and the sensitivity analysis are given further verification.The paper is divided into five chapters as follow:Chapter one, Introduction, discusses the background and meaning of the research. In this part, introduce the main content, technological route and methods.Chapter two, Knowledge of Related theory, introduces the DEA model. The part of this paper summaries the development and general application steps of DEA method research, and detailedly introduces two basic models of DEA method which are C2R model, BCC model. Following, introduces the basic concepts of sensitivity analysis.Chapter three, Sensitivity Analysis of DEA-DA method in the fuzzy environment, firstly introduces the DA model and DEA-DA model. The sensitivity analysis on the DEA-DA model with L-R fuzzy numbers (FDEA-DA model) is mostly discussed in this part of the paper. The sufficient conditions are obtained such that the discriminant result of FDEA-DA model will keep invariable when new decision making units are adding. If the discriminant result keeps invariable and the optimal solution is consistent before and after added decision making units, the conditions that the increased decision making units will satisfy are also deduced. At last, a numerical example is presented.Chapter four, Sensitivity Analysis of DEA method based on the benchmarking, firstly introduces the Benchmarking which is a new manage way, integrates the Benchmarking and DEA model, so the sensitivity analysis of DEA-Benchmarking model is discussed in this part of the paper. When the characteristic data take place perturbation, or the new decision-making unit is increased, the necessary and sufficient conditions are given, which ensure the effectiveness invariable of the unit benchmarking. At last, the conclusions of the model and the sensitivity analysis are given further verification by the 2003 China's major container port and the sample data of the first quarter of 2007 ship imports and exports of the countries.Chapter five, Conclusions and expectations of the paper, draw the main creating points of the paper and present the main direction and the field which should be studied in the future.