An optimization model based on Neural Network and Particle Swarm: an application case from the UAE

This paper addresses the combination of Neural Network (ANN) and Particle Swarm (PS) for optimization modeling. To illustrate the proposed methodology, an application case is shown to optimize the business results of a company. We estimate business results as a function of seven criteria through an ANN. The ANN is embedded in a PS metaheuristics to provide “optimal” profiles of companies based on the level of proficiency in the seven criteria. Our approach is tested using data from the quality auditors' score of 60 industrial firms in Abu Dhabi for the Sheikh Khalifa Industrial Award (SKIA) in 2000 and 2001. The choice of both algorithms (ANN and PS) is motivated by the fact that within the management system of companies, Business Result is the output of a learning process utilizing key company's variables and when competing over time, companies are evolving by auto/mutual benchmarking their performance in the market as swarm does when moving together.