A New Adaptive BFO Based on PSO for Learning Neural Network

Vol. 1  Issue 3
Year:2013
Issue:Sep-Nov
Title:A New Adaptive BFO Based on PSO for Learning Neural Network
Author Name:Ahmed Alostaz and Mohammed Alhanjouri
Synopsis:
In this paper, we introduce a new learning algorithm for neural network. A New Adaptive bacteria foraging optimization based on particle swarm optimization(ABFO_PSO) is used in learning neural network. This paper reviews Feed Forward Neural Network (FFANN), and the drawback of back- propagation learning method. Particle Swarm Optimization (PSO) is also described. Moreover, using the PSO in learning neural network is reviewed. Bacterial Foraging Optimization (BFO) is a novel heuristic algorithm inspired from forging behavior of E. coli. It is predominately used to find solutions for real-world problems, but it has problem with time and convergence behaviour. To introduce ABFO_PSO, that provide solution for BFO problem, we make a hybrid between PSO and BFO. Moreover, BFO and ABFO_PSO are applied for learning neural network. The comparison between the results of ABFO_PSO, BFO and PSO for learning neural network shows the strength of new method.