Efficiency measurement for agent simulation based on DEA with imprecise data

This paper has proposed a framework in order to analyze in conditional efficient state. Proposed model can classify the objective DMU with imprecise data into following four state: (1) efficient state in any case, (2) efficient state in the condition that objective DMU has minimum input and maximum output and other DMU has any input/output levels, (3) efficient state in the condition that objective DMU has minimum input and maximum output and other DMU has qualified input/output levels, and (4) inefficient state in any case. In addition, this paper has applied the proposed method to the data for 25 soccer teams that participated in the Japan preliminaries of Robocup 2008 and demonstrated its effectiveness.