Soft cluster-rectangle method for eliciting criteria weights in multi-criteria decision-making

Area of Science:

• Decision Sciences
• Operations Research
• Artificial Intelligence

Background:

• Subjective weighting methods in multi-criteria decision-making (MCDM) rely on decision-maker inputs.
• These methods face challenges due to input uncertainty and pairwise comparisons, impacting weight reliability.

Purpose of the Study:

• Introduce a novel MCDM method, Soft Clusters-Rectangles (SCR), to overcome limitations of existing subjective weighting techniques.
• Address uncertainty and avoid pairwise comparisons in criteria weight determination.

Main Methods:

• The SCR method utilizes fuzzy logic and analytic geometry, avoiding pairwise comparisons.
• Weights are calculated based on criteria membership values across three clusters (immaterial, mediocre, vital).
• Geometric computation involving areas of rectangles determines the final weights.

Main Results:

• The SCR method successfully determines criteria weights without pairwise comparisons.
• Results demonstrate reduced uncertainty compared to traditional subjective methods.
• Findings show similarities to objective weighting techniques and comparable results to hybrid methods.

Conclusions:

• The SCR method offers a reliable alternative for criteria weighting in MCDM.
• It effectively handles uncertainty and avoids the pitfalls of pairwise comparisons.
• The method shows promise for applications like autonomous vehicle route selection.