Toward Improving the Fuzzy KNN Algorithm Based on Takagi-Sugeno Fuzzy Inference System

In this paper we present a new approach in order to improve the performance of the Fuzzy K-Nearest Neighbor algorithm (Fuzzy KNN algorithm). We propose to use a Takagi-Sugeno Fuzzy Inference System with the Fuzzy KNN algorithm to improve classification accuracy. Also, we have used different measures to calculate the distance between the neighbors and the vector to be classified, such as the: Euclidean, Hamming, cosine similarity and city block distances. These distances represent the inputs for the Takagi-Sugeno Fuzzy Inference System. Simulation results with a classification problem show the potential of the proposed approach.