Interval efficiency improvement in DEA by using ideal points

Based on literatures in recent years on DEA models with interval data, this paper aims at enhancing the interval efficiencies for inefficient DMUs. Ideal points are used to determine the extent of inefficient DMUs that can be improved. Numbers of specific programs are established to obtain the target point as a benchmarking on the efficient frontier, and bring the adjusted inputs and outputs as respectively close to their target ones as possible. The interval efficiencies can be improved by increasing only outputs, decreasing only inputs or adjusting both simultaneously, so that the lower bound of the interval efficiency can be increased as much as possible while the upper bound can attains one. The research is based on the interval data setting, so the approach is proposed to make sure the final adjusted data are still interval data. Finally, a numerical example is shown to illustrate the idea of the method.