Performance Evaluation And Resource Allocation Under Uncertain Environment

Performance evaluation is a process by using certain evaluation methods,quantitative measures and assessment criteria to assess the degree of realization of the goal set by the centralized departments.In order to guide the operation behaviors of firms and help the managers find the gaps among different firms and improve the operation efficiency and economic efficiency,decision makers should evaluate the operation performance scientifically and reasonably.Resource allocation refers to the selection and allocation of relatively scarce resources for different purposes.Whether resource allocation is reasonable or not has a significant influence on a firm's success or failure and even a Country's economic development.In the area of management and economy,there are both theory and application meaning to study the performance evaluation and resource allocation problems.Due to the complexity and uncertainty in real production and real life,the problems of performance evaluation and resource allocation are also under uncertain environment.The two theoretical foundations of this paper are multiple attribute decision analysis(MADA)and data envelopment analysis(DEA).As an important part of decision science,MADA has been applied to various of fields such as management,economics and engineering.DEA,as a non-parametric evaluation method,it does not require weight information,so has a good advantage compared with other methods.In addition,DEA has been applied to solve the problems of resource allocation,cost allocation and production planning.The main work of this paper is organized as follows:Chapter 1 mainly introduces the concepts of performance evaluation and resource allocation,the importance of performance evaluation and resource allocation under uncertain environment,the related theories of multiple attribute decision analysis and data envelopment analysis,and briefly reviews the related literatures.Then,the main contents of this thesis are summarized.Chapter 2 develops the performance evaluation methods with uncertain weight information.First,due to evaluators tend to overestimate themselves while underestimate others in 360-degree feedback system,we develop a weight adjustment approach based on consistency analysis.In fact,the weight adjusted method is a kind of mechanism design,which aims to bound strategic evaluation or reduce deviations caused by psychological factors.Second,to address the problem of uncertain weight information in multiple attribute decision making,we develop an approach aboutdominance relation and ranking stability analysis of alternatives based on all feasible weights rather than determining a set of certain weights.Third,by combining the concept of weight space and cross evaluation,a novel multiple attribute decision making approach based on the score of mean,variance and variable coefficient is proposed.In addition,this approach is extended to the environment of uncertain attribute values.Chapter 3 is the performance evaluation based on non-homogeneous network DEA.Conventional DEA models assume that the decision making units are homogeneous.However,in our study,the assumption of homogeneity is not satisfied.In recent literatures,some studies focus on establishing one-stage non-homogeneous DEA model.In this chapter,we develop a two-stage network non-homogeneous DEA evaluation model,in which both the inputs,intermediates and outputs are non-homogeneous among DMUs.And the proposed approach is applied to evaluate the research performance of the departments in a Chinese comprehensive university.Chapter 4 concentrates on fixed cost and resource allocation by considering technology heterogeneity among DMUs.The existing approaches allocate fixed cost or resource primary based on efficiency efficiency maximized or efficiency invariance principle.Specifically,the concepts of meta-efficiency and group efficiency as well as meta-technology ratio are firstly introduced to reflect the technology level of the DMUs.Then two centralized DEA models considering technology heterogeneity are proposed to allocate fixed cost and resources,respectively.Finally,two numerical examples are presented to illustrate the feasibility and superiority of the proposed approach compared with prior studies.Chapter 5 studies the fixed cost allocation problem in a generalized two-stage structure.Existing models allocate the fixed cost by treating the DMUs as black-boxes while the internal production structures of DMUs are ignored.To open the black box,we deal with the fixed cost allocation problem for a general two-stage network production structure,in which both external inputs and outputs exist.First,we develop an additive two-stage network DEA model to evaluate the performance of DMUs with fixed cost.Then,by introducing the concept of satisfaction degree,a fixed cost allocation plan is proposed based on the principle of maximizing the fair degree.Finally,a numerical example is illustrated to verify the feasibility and superiority of the proposed approach.Chapter 6 is the carbon emission allocation approach based on gradually efficiency improvement.In order to deal with the problem of excessive carbon emission,in recent years,many organizations or national governments have taken some measures special for carbon emission reduction.Specially,China have promised in Copenhagen Climate Conference that carbon emission permits per unit of GDP(or carbon intensity)in 2020 should be reduced by 40-45% based on that in 2005 level.To achieve this goal,allocating an appropriate amount of carbon emission to each provincial administrative region fairly is a great effective way.A multiple-objective DEA programming is proposed to allocate the carbon emission permits based on the principles of gradually efficiency improvement and emission reduction planning.An algorithm is developed to obtain the global optimal solution.Some policy suggestions are given by analyzing the empirical results.