A Novel Approach to Fuzzy Implication Through Fuzzy Linear Regression
Fani I. Gkountakou, Avrilia Konguetsof, Georgios Souliotis, Basil Papadopoulos · Mathematical Modelling and Engineering Problems · 2023
Fuzzy rule-based processes have traditionally incorporated a variety of fuzzy implications using modus ponens, modus tollens, and fuzzy negations.This study introduces a novel method of fuzzy implication utilizing Fuzzy Linear Regression (FLR) with triangular fuzzy numbers.This approach was applied to evaluate the relationship between parameters influencing concrete and the compressive strength of sustainable rice husk ash (RHA) concrete.FLR, a technique for modeling relationships between inputs and outputs in a fuzzy environment, was employed to determine a fuzzy output with a specific truth value.This truth value represented the degree of truth of an entire fuzzy implication.The data used in this study were derived from real experimental results.The analysis showed that the FLR method produced accurate outputs, as indicated by a low Theil's inequality coefficient (Theil's U=0.1).The results suggest that FLR can effectively manage uncertainties in data and holds potential as an alternative method for fuzzy implication.