Noise Effects in Fuzzy Modelling Systems.pdfVIP

Noise Effects in Fuzzy Modeling Systems:Three Case Studies P.J. COSTA BRANCO and J.A. DENTE Laboratório de Mecatrónica Instituto Superior Técnico Av. Rovisco Pais 1049-011 Lisboa Codex PORTUGAL pbranco@alfa.ist.utl.pt http:\\pbranco.ist.utl.pt Abstract: - Noise is source of ambiguity for fuzzy systems. Although being an important aspect, the effects of noise in fuzzy modeling have been little investigated. This paper presents a set of tests using three well-known fuzzy modeling algorithms. These evaluate perturbations in the extracted rule-bases caused by noise polluting the learning data, and the corresponding deformations in each learned functional relation. We present results to show: 1) how these fuzzy modeling systems deal with noise; 2) how the established fuzzy model structure influences noise sensitivity of each algorithm; and 3) whose characteristics of the learning algorithms are relevant to noise attenuation. Key-Words: - fuzzy modeling, fuzzy function approximation, fuzzy systems, learning interference, noise. employing the same three fuzzy algorithms to study the 1 Introduction noise effects on fuzzy models. x One permanent problem in systems’ modeling, 2 ZE despite the use of