A New Classifier Design with Fuzzy Functions

摘要: This paper presents a new fuzzy classifier design, which constructs one for each partition of given system. The approach, namely Fuzzy Classifier Functions (FCF), is an adaptation our generic design on to classification problems. approach couples any clustering algorithm with method, in unique way. presented model derives functions (rules) from data classify patterns into number classes. c-means used capture hidden and linear or non-linear function build pattern identified. performance enhanced by using corresponding membership values the vectors as additional input variables. FCF proposed alternate representation reasoning schema rule base classifiers. method evaluated comparison experiments standard methods cross validation test patterns.