Gaussian Aggregation Operators and Applications to Intuitionistic Fuzzy Classification

Abstract In this study, we introduce novel aggregation operators for intuitionistic fuzzy values based on the Gaussian error function. We define the Gaussian triangular-norm and triangular-conorm operations using an Archimedean framework and propose the Gaussian weighted arithmetic (GWA) and the Gaussian weighted geometric (GWG) aggregation operators. These operators are applied to the classification of the Genus Iris dataset, using an improved cosine similarity measure and fuzzy classification algorithms. We demonstrate the effectiveness of these methods in handling uncertainty and improving classification accuracy. Our experimental results show that the GWA and GWG aggregation operators achieve superior performance, particularly in distinguishing between closely related species, with accuracy metrics surpassing some previous methods. This work highlights the utility of Gaussian-based fuzzy logic in complex classification tasks, offering insights into improving machine learning models dealing with imprecise data.