Abstract—The main purpose of this paper is that how to make Artificial Neural Networks (ANN) dynamic in the sense that it can decide that which architecture from given set of architecture has the optimal results. For this we need to use some optimization technique to get the optimum architecture of the ANN. In this paper Particle Swarm Optimization Technique is used. Particle Swarm Optimization (PSO) has applied to variety of optimization problems and it provides good results. In this research different architecture optimization techniques for ANN applications are reviewed. Researchers have proposed different techniques of ANN optimizations but still there required efficient ANN architecture optimization technique. In this paper the survey of ANN optimization techniques and PSO is presented and then the solution to make the ANN dynamic using PSO is proposed
Index Terms—PSO, particle, layer.
The authors are with Iqra University, Islamabad, Pakistan (email: qamar.bhk@gmail.com, jamil.ahmad@abasyn.edu.pk, waqas_bangyal@hotmail.com)
[PDF]
Cite: Qamar Abbas, Jamil Ahmad, and Waqas Haider Bangyal, "Dynamic Hidden Layers Selection of ANN Architecture Using Particle Swarm Optimization," International Journal of Engineering and Technology vol. 5, no. 2, pp.195-197, 2013.
