eprintid: 5358
rev_number: 2
eprint_status: archive
userid: 1
dir: disk0/00/00/53/58
datestamp: 2023-11-09 16:17:05
lastmod: 2023-11-09 16:17:05
status_changed: 2023-11-09 16:01:24
type: article
metadata_visibility: show
creators_name: Mustafa, M.R.
creators_name: Isa, M.H.
title: Comparative study of MLP and RBF neural networks for estimation of suspended sediments in Pari River, Perak
ispublished: pub
note: cited By 5
abstract: Estimation of suspended sediments in rivers using soft computing techniques has been extensively performed around the world since 1990's. However, accuracy in the results was always found to be highly desired and a profound crucial task. This study presents a thorough comparison between the performances of best basis function of Radial Basis Functions (RBF) and the best training algorithm in Multilayer Perceptron (MLP) neural networks for prediction of suspended sediments in Pari River, Perak, Malaysia. Time series data of water discharge and suspended sediments was used to develop MLP and RBF models. A comparison between six basis functions was performed to identify the most appropriate and best basis function for the selected time series of the river's data. The performance of the models was compared using several statistical measures including coefficient of determination, coefficient of efficiency and mean absolute error. The performance of the best RBF function was compared with the previously identified best training algorithm of MLP neural networks. The results showed that comparison of various basis functions is always advantageous to achieve the most appropriate basis function for the accurate prediction of the time series data. The results also showed that the performances of both particular RBF and MLP models were close to each other and capable to capture the exact pattern of the sediment data in the river. However, the RBF model showed some inconsistency while predicting the time series data. Furthermore, RBF modeling required more investigation to choose appropriate value for the predefined parameters as compared to MLP modeling. © Maxwell Scientific Organization, 2014.
date: 2014
publisher: Maxwell Science Publications
official_url: https://www.scopus.com/inward/record.uri?eid=2-s2.0-84901443974&doi=10.19026%2frjaset.7.740&partnerID=40&md5=4add87994de5adf6ddb25bb4d6abc386
id_number: 10.19026/rjaset.7.740
full_text_status: none
publication: Research Journal of Applied Sciences, Engineering and Technology
volume: 7
number: 18
pagerange: 3837-3841
refereed: TRUE
issn: 20407459
citation:   Mustafa, M.R. and Isa, M.H.  (2014) Comparative study of MLP and RBF neural networks for estimation of suspended sediments in Pari River, Perak.  Research Journal of Applied Sciences, Engineering and Technology, 7 (18).  pp. 3837-3841.  ISSN 20407459