Evaluation of Parameter Settings for Training Neural Networks Using Backpropagation Algorithms

Leema N Leema N,Elgin Christo V R Elgin Christo V R,Khanna H Nehemiah,Kannan A Kannan A

doi:10.4018/ijoris.2020100104

Leema N Leema N, Elgin Christo V R Elgin Christo V R + Show 2 more

Open Access

https://doi.org/10.4018/ijoris.2020100104

Copy DOI

Abstract

Artificial neural networks (ANN) are widely used for classification, and the training algorithm commonly used is the backpropagation (BP) algorithm. The major bottleneck faced in the backpropagation neural network training is in fixing the appropriate values for network parameters. The network parameters are initial weights, biases, activation function, number of hidden layers and the number of neurons per hidden layer, number of training epochs, learning rate, minimum error, and momentum term for the classification task. The objective of this work is to investigate the performance of 12 different BP algorithms with the impact of variations in network parameter values for the neural network training. The algorithms were evaluated with different training and testing samples taken from the three benchmark clinical datasets, namely, Pima Indian Diabetes (PID), Hepatitis, and Wisconsin Breast Cancer (WBC) dataset obtained from the University of California Irvine (UCI) machine learning repository.

Highlights

Researchers aim to build a computing system that will operate intelligently like a human brain
The algorithms were evaluated with different training and testing samples taken from the three benchmark clinical datasets, namely, Pima Indian Diabetes (PID), Hepatitis, and Wisconsin Breast Cancer (WBC) dataset obtained from the University of California Irvine (UCI) machine learning repository
The efficiency of the developed BP trained Artificial Neural Network (ANN) classifier has been tested with three clinical datasets namely, Wisconsin Breast Cancer, Pima Indian Diabetes and Hepatitis obtained from the UCI machine learning repository

Summary

Introduction

Researchers aim to build a computing system that will operate intelligently like a human brain. The Artificial Neural Network (ANN) facilitates the information processing in an intelligent manner (Akinyokun, 2002; Bezdek, 1993), and is inspired by the biological neural system. A biological nervous system is a large interconnection of neurons located within the brain. The functional equivalent of an artificial neuron is known as computational neuron or a node (Eluyode, Akomolafe & MNCS, 2013). These neurons are structured hierarchically by layers and interconnected between them like the biological nervous systems.

Objectives

Methods

Results

Conclusion