資料包絡分析法（DEA）已普遍的被使用在評估國營和民營公司的相對效率。傳統資料包絡分析模式是以最大化個別公司的效率為基礎：一次只考慮一個決策單位（DMU）；然而，如果要評估N 個公司的效率得分，則需要N 次類似的分解以便於達到最後的目標。過去有學者提出多目標線性規劃問題（MOLP）和資料包絡分析法之間的有趣關係，但是多目標資料包絡分析並不是簡單的問題。第一年提出了一個簡單的多目標線性規劃來求得可行解，而不是解決從原問題轉換成對偶問題的DEA 權重；因此，評估N 個決策單位的效率得分，所需的計算次數可有效的減少成一次。本研究運用評估IMD 會員國之國家競爭力的實際案例，來說明提出的研究想法。第二年承接第一年多目標的觀念，再加入模糊理論，因為在現實環境中許多情況對於觀測值的描述往往都是的不具有精確性或模糊性，並無法在客觀的標準下衡量，因此提出模糊多目標線性規劃(Fuzzy Multiple Objective Linear Programming, FMOLP)法，並以各縣市綠色環保機關作為實證的對象。承接前二年的理論，第三年再加入迴歸分析，由於當面臨必須判斷解釋變數（explanatory variable；或稱投入變數input variable）與反應變數（response variable；或稱產出變數output variable）之間的關係，迴歸分析(regression analysis) 即是解決此類問題的一種有效方法，因此發展出模糊資料包絡迴歸模式，並以物流產業做為研究對象。 The data envelopment analysis (DEA) is popularly used to evaluate the relative efficiency among public or private firms. The traditional DEA model is established on the basis of individually maximizing each firm’s efficiency: only one decision making unit (DMU) is considered at one time; thus, if there are n firms for computing efficiency scores, the resolution of n similar problems is necessary so as to achieve the final objective. Some scholars had mentioned the interesting relationship between the multi-objective linear programming (MOLP) problem and the DEA problem before, but the multi-objective fractional DEA is not simple. In the first year, we propose a simple multi-objective linear programming rather than the fractional type to resolve the DEA weights from the primal and dual perspective for a non-inferior solution set; thus, the efficiency scores of n computations of DMUs can be effectively reduced to only one. An actual example of IMD world competitiveness is presented by this interesting idea. The second year continues the Multiple Objective Linear Programming theory of the past year, and we join the previous model with the fuzzy theory. In fact, description of observed value doesn’t have imprecise and fuzziness, so it doesn’t need to measure in objectivity. We brought up an idea of Fuzzy Multiple Objective Linear Programming and apply the model to the green industry of the Environmental Protection Department in Taiwan’s Counties. Third year continues the theory of the past two years by joining the way of regression analysis, due to face must judge the relation between the parameter of explaining; (or called input parameter) and response parameter; (or called output parameter). Regression analysis is a kind of effective method to solve this kind of problem. Consequently to develop fuzzy regression and regard logistics industry as the practical case study for the research.
