On the Comparison of Fuzzy Kernel Regression Estimator and Fuzzy Radial Basis Function Networks
Autores:
Nimet YAPICI PEHLİVAN Ayşen APAYDIN
Fecha:
2010-03-30
Publicador:
Gazi University Journal of Science
Fuente:
Tipo:
Peer-reviewed Article
Tema:
No aplica
Descripción:
In this paper, we suggest two fuzzy estimators in nonparametric regression: fuzzy kernel regression (FNPR) estimator and fuzzy radial basis function (FRBF) networks. Both FNPR estimator and FRBF networks are applied to original data taken from an experiment. We obtain MSE values of the FNPR estimator and FRBF networks and then compare them. We show that the FNPR estimator is more efficient than the FRBF networks. Key Words: Fuzzy number, Fuzzy kernel regression estimator, Nonparametric regression, Neural networks, Fuzzy radial basis function networks.