VIKOR
AvailableVIseKriterijumska Optimizacija I Kompromisno Resenje. Opricovic (1998)
The idea
- No acceptable advantage: A7 leads A1 by only 0.0018 in Q, less than DQ = 0.1111.
- Acceptable stability: A7 is also first by R.
Compromise set: A7, A1, A8
Below the dashed line Q stays within DQ of the leader; without an acceptable advantage those alternatives form the compromise set.
Show as table
| Alternative | S | R | Q | In the compromise |
|---|---|---|---|---|
| A7 | 0.1447 | 0.0350 | 0.0270 | yes |
| A1 | 0.1008 | 0.0473 | 0.0287 | yes |
| A8 | 0.1752 | 0.0500 | 0.0806 | yes |
| A9 | 0.2384 | 0.1029 | 0.2423 | no |
| A2 | 0.3507 | 0.1758 | 0.4807 | no |
| A5 | 0.5971 | 0.1333 | 0.5331 | no |
| A3 | 0.6187 | 0.1338 | 0.5473 | no |
| A6 | 0.6953 | 0.1681 | 0.6743 | no |
| A4 | 0.9160 | 0.2140 | 0.9163 | no |
| A10 | 0.8299 | 0.2500 | 0.9472 | no |
Data: Ten alternatives, seven criteria (Keshavarz Ghorabaee et al. 2015).
VIKOR picks the compromise that is best overall (S) while keeping the worst single-criterion regret small (R).
When to use it
- Conflicting criteria and decision makers who want a compromise.
- An explicit test of whether the winner is stable (C1, C2).
- Units may change: linear normalization is unaffected, unlike TOPSIS.
Algorithm
1Best and worst per criterion
2Group utility and individual regret
3Compromise index
S* = min S, S- = max S and the same for R; rank by S, R and Q ascending, lower is better.
4Acceptable advantage and stability
With a' first and a'' second by Q, and C2 meaning a' also leads by S or R, the compromise set is a' alone if both hold, a' and a'' if only C1 holds, and every alternative within DQ of a' if C1 fails.
Used together with
Why these pairs
- AHP then VIKOR, and fuzzy Delphi with AHP then VIKOR, in sustainability and renewable-energy papers (Mardani et al. 2016). 1,753 AHP-VIKOR co-mentions in OpenAlex.
- VIKOR together with TOPSIS for comparison (Opricovic & Tzeng 2004).
- BWM then VIKOR, for example Gupta (2018).
Reference example
Details
Opricovic, S.; Tzeng, G.-H. (2004). Compromise solution by MCDM methods: A comparative analysis of VIKOR and TOPSIS. European Journal of Operational Research 156(2), 445-455. doi:10.1016/S0377-2217(03)00020-1
Published example (Tables 1-3), reproduced by the tests: in the mountain-climbing example S, R and Q match Table 3 exactly, C1 and C2 hold, the compromise is A2, and the other unit encoding (problem φ) gives the same numbers. The Q ranks also equal the published VIKOR ranks of Keshavarz Ghorabaee et al. (2015) in 7 of 7 weight sets with v = 0.5 (the paper does not state v).
Checked by this site's test suite.
Sources
Show 6 sources
- Opricovic (1998). Multicriteria Optimization of Civil Engineering Systems. Faculty of Civil Engineering, Belgrade. Book in Serbian; the English statement used here is Opricovic & Tzeng (2004).
- Opricovic (1998), Multicriteria Optimization of Civil Engineering Systems, Belgrade
- Opricovic & Tzeng (2004), European Journal of Operational Research 156(2), 445-455 doi:10.1016/S0377-2217(03)00020-1
- Keshavarz Ghorabaee, Zavadskas, Olfat & Turskis (2015), Informatica 26(3), 435-451 doi:10.15388/Informatica.2015.57
- Mardani et al. (2016), VIKOR Technique: A Systematic Review, Sustainability 8(1), 37 doi:10.3390/su8010037
- pyDecision 5.1.1, vikor_method github.com/Valdecy/pyDecision