Nonlinear function approximation using radial basis function neural networks

H. Husain1, M. Khalid1, R. Yusof1
1Center For Artificial Intelligence and Robotics (CAIRO), Faculty of Electrical Engineering, Universiti Teknologi Malaysia, Kuala Lumpur, Malaysia

Tóm tắt

Radial basis function neural networks (RBFNN) which are best suited for nonlinear function approximation, have been successfully applied to a wide range of areas including system modeling. The two-stage training procedure adapted in numerous RBFNN applications usually provides satisfactory network performance. Though this method is proven to allow faster training and improves convergence, the initial stage of selecting the network centers pose a problem of creating a larger architecture than what is required. This limitation holds true in applications with large data samples. Various techniques have been developed to choose a sufficient number of centers to suit the network structure. Orthogonal least squares and input clustering are two of such methods that show considerable results of which can provide an amicable solution to the above problem. This paper presents a comparative study on the performance achieved by the two techniques demonstrated when applying the RBFNN in modeling of nonlinear functions and an investigation based on their capabilities in handling over-parameterization problems.

Từ khóa

#Function approximation #Radial basis function networks #Least squares methods #Least squares approximation #Clustering methods #Clustering algorithms #Intelligent networks #Modeling #Convergence #Multi-layer neural network

Tài liệu tham khảo

lee, 1999, Robust Radial Basis Function Neural Network, IEEE Trans Systems Man and Cybernetics-Part B Cybernetics, 29, 674, 10.1109/3477.809023 10.1109/ICSMC.1998.728118 chang, 1997, Recursive Orthogonal Least Square Method and Its Application in Non-Linear Adaptive Filtering, 1392 10.1109/72.839002 fun, 1999, Recursive Orthogonal Least Square Learning with Automatic. Weight Selection for Gaussian Neural Networks, 1496 10.1109/72.80341 haykin, 1999, Neural Networks: A Comprehensive Foundation 10.1109/ACC.2000.876739 10.1049/ip-f-2.1992.0054 10.1109/72.977289 10.1109/72.857766 lucks, 1999, A Radial Basis Network for Function Approximation, 1099 10.1109/ICPR.2000.906151 10.1109/IJCNN.1999.831151