Multi-Kernel Fusion for RBF Neural Networks

TL;DR

This paper introduces a novel multi-kernel RBF neural network where each kernel has its own weight, leading to improved convergence, better local minima, and resilience, demonstrated across classification, system identification, and function approximation tasks.

Contribution

It proposes a new multi-kernel RBFNN with individual kernel weights, enhancing performance and robustness over existing convex combination methods.

Findings

Faster convergence rate observed in experiments.

Improved performance in local minima avoidance.

Superior results compared to state-of-the-art methods.

Abstract

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.