Title: An improved functional link neural network learning using artificial bee colony optimisation for time series prediction

Authors: Yana Mazwin Mohmad Hassim; Rozaida Ghazali

Addresses: Faculty of Computer Science and Information Technology, Universiti Tun Hussein Onn Malaysia, Batu Pahat, Johor 86400, Malaysia ' Faculty of Computer Science and Information Technology, Universiti Tun Hussein Onn Malaysia, Batu Pahat, Johor 86400, Malaysia

Abstract: Functional link neural network (FLNN) has emerged as an important tool used for function approximation and IT application on physical time series prediction. The standard learning scheme used for the training of FLNN is the Backpropagation (BP) learning algorithm. However, one of the crucial problems with BP learning algorithm is it tends to easily get trapped on local minima and thus affect the performance of FLNN. This paper proposed an alternative learning scheme for FLNN by using an artificial bee colony (ABC) optimisation algorithm as an attempt to overcome this problem. The performance of FLNN-ABC model is measured based on the prediction task on the physical time series data. The result of the prediction made by FLNN-ABC is compared with the original FLNN architecture and towards the end we found that FLNN-ABC gives better result in predicting the next-day ahead prediction.

Keywords: functional link neural networks; FLNN; artificial bee colony; ABC; time series prediction; optimisation.

DOI: 10.1504/IJBIDM.2013.059264

International Journal of Business Intelligence and Data Mining, 2013 Vol.8 No.4, pp.307 - 318

Published online: 28 Jun 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article