Binary Particle Swarm Optimization Based Algorithm for Feature Subset Selection

The feature subset selection can be considered as a global combinatorial optimization problem in which the optimum subset of features is selected from a large set of features. Lots of techniques have developed so far, still research is going on to find better solution in terms of optimality and computational ease. In this work an algorithm based on binary particle swarm optimization (bPSO) is proposed for feature subset selection. From simple simulation experiments it has been found that bPSO based algorithm performs well and computationally less demanding than genetic algorithm, another population based evolutionary search technique.