Mathematical Models of Experience-Based and Dynamic Experience-Based Fuzzy Classification

This paper presents novel mathematical models to decide similarity functions for experience-based and dynamic experience-based fuzzy classification. By extending crisp partition to fuzzy partition and introducing statistical approach, we firstly establish models for experience-based fuzzy classification. Based on these models, we propose a new mathematical model for dynamic experience-based fuzzy classification. The new model is more practical than previous ones. We illustrate how to utilize the new model with an example of classifying engineering materials.