This page is published in English.
DEA
HFPE - Hesitant Fuzzy Peer-Evaluation (benevolent / aggressive cross-efficiency, Zhou-Chen-Xu-Meng 2018)
Cross-efficiency DEA extended to HFS - each alternative is evaluated by both its own optimal HFEA weights (self-evaluation, score E_{ee}) and by the optimal weights of every other alternative (peer evaluation, cross-efficiency E_{el}). Benevolent strategy maximises peer scores; aggressive strategy minimises them. Final ranking by column-mean of cross-efficiency matrix.
Zhou, W., Chen, J., Xu, Z. S., Meng, S.2018doi:10.1016/j.ins.2018.07.002 ↗
Overview
HFPE produces full discrimination where HFEA ties at m_e=1. The cross-efficiency column mean ĒL is a peer-reviewed score: a high ĒL means the alternative performs well not just under its own most favourable weights, but also under the weights preferred by its competitors. Benevolent strategy is more generous; aggressive is more discriminating. Choose based on decision context.
- Data
- Hesitant
Edge cases and pitfalls
Non-uniqueness of LP optimal weights: different LP solvers or feasibility tolerances may produce different cross-efficiency matrices for the same data.
Benevolent and aggressive strategies can give very different rankings - always report which strategy was used.
Computational cost scales as K² LP solves; for large K (> 50), consider approximation or sampling strategies.
Works with
Its derived weights can feed
How to cite
Zhou, W.; Chen, J.; Xu, Z. S.; Meng, S. (2018). Hesitant fuzzy preference envelopment analysis and alternative improvement. Information Sciences. https://doi.org/10.1016/j.ins.2018.07.002
System ID, as it appears in reports and the API
HFPE