Comparative Analysis of Methods for Determining Number of Hidden Neurons in Artificial Neural Network

Abstract

Neurons in an artificial neural network are grouped in three layers: input, output and hidden layer. Determination of an optimal number of neurons in hidden layer is one of the major difficulties in the process of creating artificial neural network topology. The main goal of this paper is to explore and compare existing methods for determining number of hidden neurons. The research is conducted on two separate datasets with different number of input values and different number of training pairs.