Discrete Choquet Integral Based on Quasi-Overlap Functions on Intuitionistic Fuzzy Sets and Its Application in Multi-Class Ensemble Learning

As a extension of fuzzy sets, intuitionistic fuzzy sets play an essential part in processing uncertain information. Quasi-Overlap functions have emerged in recent years as a tool for information aggregation. The combination of intuitionistic fuzzy sets with quasi-overlap functions offers new methodologies for addressing uncertain information. This study first introduces the notion of quasi-overlap functions on intuitionistic fuzzy sets and presents specific examples. It then explores the generation methods for quasi-overlap functions within the framework of intuitionistic fuzzy sets. Subsequently, a novel class of discrete Choquet integral fusion operators is constructed using the generated quasi-overlap functions. Finally, the newly constructed discrete Choquet integral fusion operator is applied to the multi-class ensemble problems, resulting in the development of a new multi-class ensemble algorithm. The superiority of the proposed method is empirically validated using UCI datasets.