Feature selection with Intelligent Dynamic Swarm and Rough Set

Data mining is the most commonly used name to solve problems by analyzing data already present in databases. Feature selection is an important problem in the emerging field of data mining which is aimed at finding a small set of rules from the training data set with predetermined targets. Many approaches, methods and goals including Genetic Algorithms (GA) and swarm-based approaches have been tried out for feature selection in order to these goals. Furthermore, a new technique which named Particle Swarm Optimization (PSO) has been proved to be competitive with GA in several tasks, mainly in optimization areas. However, there are some shortcomings in PSO such as premature convergence. To overcome these, we propose a new evolutionary algorithm called Intelligent Dynamic Swarm (IDS) that is a modified Particle Swarm Optimization. Experimental results states competitive performance of IDS. Due to less computing for swarm generation, averagely IDS is over 30% faster than traditional PSO.