Comparison of sigmoidal FFANN training algorithms for function approximation problems
M. P. S. Bhatia, Veenu Veenu, Pravin Chandra · International Conference on Computing for Sustainable Global Development · 2015
The estimation of unknown function from a number of data inputs has number of various applications like in Engineering, Artificial intelligence, Statistics, Artificial Neural Networks, Genetic algorithms etc. Many papers have described the individual methods. But very less is known about the comparative performance of various methods. In this paper we give the comparative performance of the neural network using ten different approximation functions and twelve various training algorithms. Our study uses MATLAB 2013a 8.1 Neural Network toolbox for experimentation. The performance of the method on the neural network depends on the approximation function type and the various properties of training data. We found that Bayesian Regulation Backpropagation method proved to be best in performance using function 6 given in the paper out of twelve different algorithms used.