Improving Nearest Neighbor Classification Using Particle Swarm Optimization with Novel Fitness Function

A new method of feature selection is presented in this paper. The proposed idea uses Particle Swarm Optimization (PSO) with fitness function in order to assign higher weights to informative features while noisy irrelevant features are given low weights. The measure of Area Under the receiver operating characteristics Curve (AUC) is used as the fitness function of the particles. Experimental results claim that the PSO-based feature weighting can improve the classification performance of the k-NN algorithm in comparison with the other important method in realm of feature weighting such as Mutual Information, Genetic Algorithm, Tabu Search and chi-squared (X-2). Additionally, on synthetic data sets, this method is able to allocate very low weight to the noisy irrelevant features which may be considered as the eliminated features from the data set.