Ranking
Criteria Removal: Sensitivity analysis by sequential criterion exclusion
Saltelli, A., Tarantola, S., Campolongo, F., Ratto, M. · 2004
Overview
Robustness diagnostic: criterion removal sensitivity. Output typically robustness_score (higher value = preferred).
Strengths
- •Method-specific: Robustness diagnostic: criterion removal sensitivity
Limitations
- •Assumes: A base ranking method is selected
- •Assumes: Computational budget for repeated runs
Method assistant
Grounded explanations: it explains the method, it does not compute.
Assumptions to verify
- •A base ranking method is selected
- •Computational budget for repeated runs
When not to use
- •Quick analysis needed → defer robustness check
Edge cases
- •See F.steps and D.parameters for CRITERIA-REMOVAL-specific edge handling. recommendation_metadata.assumptions_to_verify also lists boundary conditions.
Common pitfalls
- •Hatalı: 'CRITERIA-REMOVAL bu varsayımı kontrol etmeden uygulamak'. Doğrusu: A base ranking method is selected
- •Hatalı: 'CRITERIA-REMOVAL bu varsayımı kontrol etmeden uygulamak'. Doğrusu: Computational budget for repeated runs
- •Hatalı: CRITERIA-REMOVAL'yi 'Quick analysis needed → defer robustness check' durumunda kullanmak: recommendation_metadata.not_recommended_when alternatif öneriyor.
Worked example
- 1.Adım 1 (F1): Validate inputs for criteria-removal. Formül: \text{Re-run base MCDM with criterion }j\text{ excluded; report }\rho_s(\text{base},\text{reduced}) Anchor: Saltelli 2004, (pending PDF page verification)
How to cite
Saltelli, A.; Tarantola, S.; Campolongo, F.; Ratto, M. (2004). Sensitivity Analysis in Practice. Wiley, Chichester. https://doi.org/10.1002/0470870958