Rohrer’s Index Prediction Using Neural Network.

OLAIJU, O.A and Are, Stephen Olusegun and OJUAWO, Olutayo (2018) Rohrer’s Index Prediction Using Neural Network. The Ilaro Journal of Management, Arts, Science and Technology (TIJMAST)., 3 (1). ISSN 2672 - 4537

Text Rohrer1 2018.pdf Download (863kB)

Abstract

Artificial neural networks can be considered effective in making predictions in which traditional methods and statistics are not suitable. In this article, by using two – layer feed forward network with tan-sigmoid transmission function in input and output layer. We can anticipate the prediction of Rohrer’s index an anthropometric statistic which combines the height and weight of an individual into a singular metric used to classify individuals into the following categories: severely underweight, underweight, normal, overweight, and obese. The authors compared different artificial neural networks architectures with the traditional multiple regression model. All of the models means are relatively close to zero. However, the breakout occurs with standard deviation. The larger the standard deviation, the greater the range of error, so ANN10 model performed best in predicting Rohrer Index.

Item Type:	Article
Subjects:	Q Science > Q Science (General)
Divisions:	Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science
Depositing User:	Mr Taiwo Egbeyemi
Date Deposited:	11 Jun 2020 15:40
Last Modified:	11 Jun 2020 15:40
URI:	http://eprints.federalpolyilaro.edu.ng/id/eprint/515

Actions (login required)

View Item