Data-Based Fuzzy TOPSIS for Alternative Ranking

Abstract

Technique for order preference by similarity (TOPSIS) solves multi-criteria decision making (MCDM) by ranking the alternatives. When the attributes are not deterministic, a Fuzzy TOPSIS method is applied. The traditional fuzzy TOPSIS depends on decision makers to determine alternative's value which considered subjective. A new method named data-based fuzzy TOPSIS proposed to diminish the dependency to decision maker. The proposed algorithm use data to determine alternative's values objectively. Subtractive Clustering (SC) and Fuzzy C-Mean (FCM) selected to transform crisp value data to fuzzy value data. Some modification applied to SC and FCM to obtain fuzzy triangular value needed by fuzzy TOPSIS.