Methods
26 of 26 methods compute in the browser and are checked against published examples.
Available
Weighting
- CRITICWeights from contrast and conflict in the data.
- EntropyWeights from how unevenly the data spread.
- SDWeights from how widely the data spread.
- MERECWeights from what removing a criterion changes.
- LOPCOWWeights from the log ratio of root-mean-square to spread.
- CILOSWeights from the relative impact loss when each criterion's best alternative is chosen.
- Equal weightsEvery criterion gets 1/n.
- BWMWeights from your answers to a few comparisons.
- AHPWeights from comparing every pair of criteria.
- ROCWeights from the importance rank order of criteria.
- SWARAChains step-down importance ratios between ordered criteria into weights.
Ranking
- TOPSISRanks by distance to the best and the worst.
- SAWAdds up weighted, rescaled values.
- VIKORBalances the best total against the worst regret.
- EDASMeasures the distance above and below the average.
- WASPASBlends a weighted sum with a weighted product.
- COPRASAdds benefit shares and a bonus for a small cost share.
- MARCOSMeasures against an imaginary ideal and anti-ideal.
- CoCoSoThree readings of a sum and a power sum, combined.
- MABACDistance above or below a border per criterion.
- MOORABenefit ratios minus cost ratios after vector normalization.
- ARASMeasures each weighted share total against an optimal row.
- CODASCombines Euclidean and Taxicab distances from the negative-ideal across pairs.
- PROMETHEE IIPairwise preference functions and leaving minus entering outranking flows.
- ELECTRE IMajority concordance and bounded discordance veto to select the outranking kernel.
- ELECTRE IIIPseudo-criteria with indifference, preference and veto thresholds, fuzzy credibility and distillation.