An improving completely ranking approach for interval data envelopment analysis

The efficiency score of decision making units (DMU) in interval DEA is interval number, traditional research on interval DEA can only divide the DMUs into three types (efficient, part efficient, inefficient) using interval efficiency score, but there are less research on the ranking of DMUs in interval DEA. This paper ranked the DMUs containing the interval number in the input and output based on the likelihood degree of the comparison between two interval numbers through ranking vector, taking into account the situation that interval number degenerates to the real number. In order to rank all the DMUs completely, we rank the efficient DMUs through the extensions of super-efficiency DEA to interval case. In the end, an empirical example is illustrated.