Intelligent DEA-based Decision Support System

Wang Chen

Abstract

Data envelopment analysis (DEA) is a method that measures the relative efficiency of a group of homogeneous decision-making units (DMUs) to determine the best practice (or effective) frontier in the presence of multiple inputs and outputs. Although there are many kinds of DEA software to help users solve the problem of calculation, many DEA software such as DEAP and DEA solver, calculate on the assumption that there is no error with the default data. If the data set such as negative value, non-discretionary indicator, ordinal indicator, and so on are used as a normal indicator, the results will be inaccurate. Since 1978, a large number of DEA models for different situations have been developed, including DEA models considering negative data, but the degree of integration and intelligence of such models is low. To avoid the calculation error of the DEA model caused by improper data application, this thesis takes the intelligent selection of the DEA model as the research object, deeply investigates the relationship between data type and DEA model, and designs and develops a data-oriented intelligent DEA decision support system based on the knowledge base, DEA model base, model matching, and other technologies. The work includes (1) To meet the calculation requirements of different data sets, an intelligent DEA model base is analyzed and designed (2) To improve the intelligence of the system, the function of DEA model selection based on knowledge is designed and implemented (3) Based on the system design scheme, the construction of an intelligent decision support system based on DEA is completed, and the functions of the system are realized. .