TY - JOUR
T1 - Multi-Kernel Fusion for RBF Neural Networks
AU - Atif, Syed Muhammad
AU - Khan, Shujaat
AU - Naseem, Imran
AU - Togneri, Roberto
AU - Bennamoun, Mohammed
PY - 2023/4
Y1 - 2023/4
N2 - A simple yet effective architectural design of radial basis function neural networks (RBFNN) makes them amongst the most popular conventional neural networks. The current generation of radial basis function neural network is equipped with multiple kernels which provide significant performance benefits compared to the previous generation using only a single kernel. In existing multi-kernel RBF algorithms, multi-kernel is formed by the convex combination of the base/primary kernels. In this paper, we propose a novel multi-kernel RBFNN in which every base kernel has its own (local) weight. This novel flexibility in the network provides better performance such as faster convergence rate, better local minima and resilience against stucking in poor local minima. These performance gains are achieved at a competitive computational complexity compared to the contemporary multi-kernel RBF algorithms. The proposed algorithm is thoroughly analysed for performance gain using mathematical and graphical illustrations and also evaluated on three different types of problems namely: (i) pattern classification, (ii) system identification and (iii) function approximation. Empirical results clearly show the superiority of the proposed algorithm compared to the existing state-of-the-art multi-kernel approaches.
AB - A simple yet effective architectural design of radial basis function neural networks (RBFNN) makes them amongst the most popular conventional neural networks. The current generation of radial basis function neural network is equipped with multiple kernels which provide significant performance benefits compared to the previous generation using only a single kernel. In existing multi-kernel RBF algorithms, multi-kernel is formed by the convex combination of the base/primary kernels. In this paper, we propose a novel multi-kernel RBFNN in which every base kernel has its own (local) weight. This novel flexibility in the network provides better performance such as faster convergence rate, better local minima and resilience against stucking in poor local minima. These performance gains are achieved at a competitive computational complexity compared to the contemporary multi-kernel RBF algorithms. The proposed algorithm is thoroughly analysed for performance gain using mathematical and graphical illustrations and also evaluated on three different types of problems namely: (i) pattern classification, (ii) system identification and (iii) function approximation. Empirical results clearly show the superiority of the proposed algorithm compared to the existing state-of-the-art multi-kernel approaches.
KW - Cosine distance
KW - Euclidean distance
KW - Function approximation
KW - Gaussian kernel
KW - Kernel fusion
KW - Neural networks
KW - Non-linear system identification
KW - Pattern classification
KW - Radial basis function
KW - Support vector machine
UR - http://www.scopus.com/inward/record.url?scp=85132813302&partnerID=8YFLogxK
U2 - 10.1007/s11063-022-10925-3
DO - 10.1007/s11063-022-10925-3
M3 - Article
AN - SCOPUS:85132813302
SN - 1370-4621
VL - 55
SP - 1045
EP - 1069
JO - Neural Processing Letters
JF - Neural Processing Letters
IS - 2
ER -