Performance assessment of fuzzy clustering models applied to urban airborne pollution

Abstract

An application in modeling a non-lineal system between temperature, humidity and urban airborne air pollution is presented. In this contribution, the implementation of cluster estimation method as a basis of a fuzzy model identification algorithm has been developed. Fuzzy clustering allowed partitioning this complex non-linear system into many linear sub-systems. Finally, comparison of the performance between two different clustering techniques for this particular case study is presented: Fuzzy C-means Clustering and Fuzzy Subtractive Clustering.