ARAS
AvailableAdditive Ratio ASsessment. Zavadskas and Turskis (2010)
The idea
The dashed row is A₀: the best value of every criterion, normalized in the same m + 1 column totals. Each bar is S, read against S₀; K = S / S₀ on the right.
Show as table
| Alternative | Optimality function S | Utility degree K |
|---|---|---|
| Optimal A₀ | 0.0997 | 1.0000 |
| ANA | 0.0980 | 0.9838 |
| Hainan | 0.0904 | 0.9068 |
| Qantas | 0.0881 | 0.8841 |
| THY | 0.0859 | 0.8619 |
| China Southern | 0.0841 | 0.8439 |
| Emirates | 0.0832 | 0.8349 |
| Air France | 0.0825 | 0.8281 |
| Lufthansa | 0.0824 | 0.8266 |
| British Airways | 0.0747 | 0.7498 |
| Air China | 0.0655 | 0.6576 |
| LATAM | 0.0655 | 0.6575 |
Data: Airlines (Bakır and Atalık 2018).
ARAS adds an optimal benchmark row built from the best value on each criterion, totals every row's weighted shares, and scores each alternative as its fraction of that optimal total.
When to use it
- All values are positive and you want a share-of-optimum score.
- You want results comparable with the Entropy, AHP or SWARA ARAS literature.
Algorithm
1Add the optimal benchmark row
Row 0 takes the column best on each criterion (maximum for benefit, minimum for cost), or an external target row when given.
2Invert cost criteria and normalize each extended column
The column sum in the denominator runs from row 0 through row m, so the optimal row takes part of each column total.
3Weight the normalized matrix
4Optimality function per row
Row 0 gives the optimal benchmark total S0, and rows 1 to m give each alternative total Si.
5Utility degree and ranking
Each alternative is scored as its share of the optimal benchmark S0 and ranked from highest K to lowest.
Used together with
Why these pairs
- Bakır & Atalık (2018) weight seven airline service-quality criteria with Entropy and rank eleven airlines with ARAS.
- Aytaç Adalı & Tuş Işık (2016) pair AHP weights with ARAS and COPRAS; the 2010 origin paper uses expert weights from 38 specialists.
- Adding row 0 to each column sum scales criterion j by 1 / (1 + p0j), so ARAS can rank alternatives differently from SAW.
Reference example
Details
Bakır, M.; Atalık, Ö. (2018). Entropi ve Aras Yöntemleriyle Havayolu İşletmelerinde Hizmet Kalitesinin Değerlendirilmesi. İşletme Araştırmaları Dergisi 10(1), 617-638. doi:10.20491/isarder.2018.410
Published example, reproduced by the tests: eleven airlines on seven Skytrax service-quality criteria with Entropy weights (Tables 3, 7, 8-11). S0 = 0.099651, all eleven Si and Ki values in Table 11 match within 0.00002, and all eleven ranks match; our Entropy also reproduces Table 7's weights. (Tables 9 and 10 accidentally reprint the A0 row on the THY line, while Table 11 uses THY's true row.) Also tested: Zavadskas & Turskis (2010, Tables 1-3, fourteen office rooms on six criteria), where Table 2 entered room 8's illumination as 390 lx (room 1) instead of 400 lx, swapping rooms 8 and 9; with the Table 1 data, room 8 ranks first under both the paper's target norms (K8 = 0.7762) and the column-best benchmark (K8 = 0.9485).
Checked by this site's test suite.
Sources
Show 6 sources
- Zavadskas and Turskis (2010). A new additive ratio assessment (ARAS) method in multicriteria decision-making. Technological and Economic Development of Economy 16(2), 159-172. doi:10.3846/tede.2010.10
- Zavadskas, E.K. & Turskis, Z. (2010), Technological and Economic Development of Economy 16(2), 159-172 (Tables 1-3) doi:10.3846/tede.2010.10
- Bakır, M. & Atalık, Ö. (2018), İşletme Araştırmaları Dergisi 10(1), 617-638 (Eqs. 1-11, Tables 3-11) doi:10.20491/isarder.2018.410
- Aytaç Adalı, E. & Tuş Işık, A. (2016), Manas Journal of Social Studies 5(2), 124-138 (Eqs. 9-15, Tables 1, 5-8) dergipark.org.tr/en/download/article-file/577434
- Keshavarz Ghorabaee, M. et al. (2016), CODAS, Economic Computation and Economic Cybernetics Studies and Research 50(3), 25-44 (Example 2) ideas.repec.org/a/cys/ecocyb/v50y2016i3p25-44.html
- pyDecision, aras_method github.com/Valdecy/pyDecision