Classification and forecasting of sustainable-resilience suppliers via developing a novel fuzzy MIP model and DEA in the presence of zero data

This study suggests a novel fuzzy super-efficiency data envelopment analysis (FS-DEA) and fuzzy mixed integer programming (F-MIP) for suppliers’ complete ranking and classification regarding sustainability and resilience paradigms. The introduced approach applies FS-DEA to estimate efficiency scores and classify suppliers into efficient and inefficient groups, given their efficiency scores. Then, it employs a two-step F-MIP model to forecast the group membership of the new supplier. The computational process of the two-step F-MIP involves identifying the misclassification and overlap in the first step and managing the overlap in the second step. The suggested approach has the following features, which cannot be found in the traditional use of DEA in the supplier selection context. First, the proposed FS-DEA model can evaluate the performance of suppliers and then yield a full ranking given the zero data. Second, the proposed FS-DEA can classify suppliers into efficient and inefficient groups given deterministic and fuzzy criteria for any level $$\alpha \in (0 1]$$ α ∈ ( 01 ] . Third, the proposed FS-DEA uses input saving index and output surplus index to have a feasible solution even when there are non-negative data. Fourth, the proposed F-MIP model minimizes the number of wrong-classified suppliers in the fuzzy context. The developed models rank and classify suppliers of the largest automobile companies in Iran Finally, a sensitivity analysis verifies the validity of the proposed F-MIP model.