PSO and DE based novel quantum inspired automatic clustering techniques

Clustering, a well-known technique, is used to divide a data set into number of groups, called clusters. Differential evolution and particle swarm optimization are robust, fast and very effective search techniques. To increase computational capability, two different quantum inspired meta-heuristics for automatic clustering, have been proposed here. An application of quantum inspired techniques has been demonstrated for automatic clustering of image data sets. These techniques are able to find optimal number of clusters “on the run” for an image data sets. As the comparative research, a comparison has been made between the proposed techniques and their conventional counterparts for four images data set. Effectiveness of the proposed techniques has been exhibited against the fitness value, standard deviation and mean of the fitness, standard error and computational time. Finally, two separate statistical superiority test, referred to as i-test and Friedman test have been performed to prove the superiority the of proposed approaches in their favor.