Cover Image

WCBP: A new water cycle based back propagation algorithm for data classification

Water Cycle algorithm is a modern nature inspired meta-heuristic algorithm to provide derivative-free solution to optimize complex problems. The back-propagation neural network (BPNN) algorithm performs well on many complex data types but it possess the problem of network stagnancy and local minima....

Full description

Saved in:
Bibliographic Details
Main Authors: Mohd. Nawi, Nazri, Khan, Abdullah, Firdaus, Naim, M. Z., Rehman, Siming, Insaf Ali
Format: Article
Language:English
Published: Asian Research Publishing Network (ARPN) 2016
Subjects:
QA299.6-433 Analysis
Online Access:http://eprints.uthm.edu.my/4299/1/AJ%202016%20%2836%29.pdf
http://eprints.uthm.edu.my/4299/
Tags: Add Tag
No Tags, Be the first to tag this record!
id my.uthm.eprints.4299
record_format eprints
spelling my.uthm.eprints.42992021-12-02T02:40:49Z http://eprints.uthm.edu.my/4299/ WCBP: A new water cycle based back propagation algorithm for data classification Mohd. Nawi, Nazri Khan, Abdullah Firdaus, Naim M. Z., Rehman Siming, Insaf Ali QA299.6-433 Analysis Water Cycle algorithm is a modern nature inspired meta-heuristic algorithm to provide derivative-free solution to optimize complex problems. The back-propagation neural network (BPNN) algorithm performs well on many complex data types but it possess the problem of network stagnancy and local minima. Therefore, this paper proposed the use of WC algorithm in combination with Back-Propagation neural network (BPNN) algorithm to solve the local minima problem in gradient descent trajectory. The performance of the proposed Water Cycle based Back-Propagation (WCBP) algorithm is compared with the conventional BPNN, ABC-BP and ABC-LM algorithms on selected benchmark classification problems from UCI Machine Learning Repository. The simulation results show that the BPNN training process is highly enhanced when combined with WC algorithm. Asian Research Publishing Network (ARPN) 2016 Article PeerReviewed text en http://eprints.uthm.edu.my/4299/1/AJ%202016%20%2836%29.pdf Mohd. Nawi, Nazri and Khan, Abdullah and Firdaus, Naim and M. Z., Rehman and Siming, Insaf Ali (2016) WCBP: A new water cycle based back propagation algorithm for data classification. ARPN Journal of Engineering and Applied Sciences, 11 (24). pp. 14132-14135. ISSN 1819-6608
institution Universiti Tun Hussein Onn Malaysia
building UTHM Library
collection Institutional Repository
continent Asia
country Malaysia
content_provider Universiti Tun Hussein Onn Malaysia
content_source UTHM Institutional Repository
url_provider http://eprints.uthm.edu.my/
language English
topic QA299.6-433 Analysis
spellingShingle QA299.6-433 Analysis
Mohd. Nawi, Nazri
Khan, Abdullah
Firdaus, Naim
M. Z., Rehman
Siming, Insaf Ali
WCBP: A new water cycle based back propagation algorithm for data classification
description Water Cycle algorithm is a modern nature inspired meta-heuristic algorithm to provide derivative-free solution to optimize complex problems. The back-propagation neural network (BPNN) algorithm performs well on many complex data types but it possess the problem of network stagnancy and local minima. Therefore, this paper proposed the use of WC algorithm in combination with Back-Propagation neural network (BPNN) algorithm to solve the local minima problem in gradient descent trajectory. The performance of the proposed Water Cycle based Back-Propagation (WCBP) algorithm is compared with the conventional BPNN, ABC-BP and ABC-LM algorithms on selected benchmark classification problems from UCI Machine Learning Repository. The simulation results show that the BPNN training process is highly enhanced when combined with WC algorithm.
format Article
author Mohd. Nawi, Nazri
Khan, Abdullah
Firdaus, Naim
M. Z., Rehman
Siming, Insaf Ali
author_facet Mohd. Nawi, Nazri
Khan, Abdullah
Firdaus, Naim
M. Z., Rehman
Siming, Insaf Ali
author_sort Mohd. Nawi, Nazri
title WCBP: A new water cycle based back propagation algorithm for data classification
title_short WCBP: A new water cycle based back propagation algorithm for data classification
title_full WCBP: A new water cycle based back propagation algorithm for data classification
title_fullStr WCBP: A new water cycle based back propagation algorithm for data classification
title_full_unstemmed WCBP: A new water cycle based back propagation algorithm for data classification
title_sort wcbp: a new water cycle based back propagation algorithm for data classification
publisher Asian Research Publishing Network (ARPN)
publishDate 2016
url http://eprints.uthm.edu.my/4299/1/AJ%202016%20%2836%29.pdf
http://eprints.uthm.edu.my/4299/
_version_ 1738581233308270592
score 13.244745

Similar Items