No abstract

Abstract. The paper presents a comparative analysis of K-means and PSO based clustering performances for text datasets. The dimensionality reduction techniques like Stop word removal, Brill's tagger algorithm and mean Tf-Idf are used while reducing the size of dimension for clustering. The results reveal that PSO based approaches find better solution compared to K-means due to its ability to evaluate many cluster centroids simultaneously in any given time unlike K-means.

The paper presents a novel approach of clustering image datasets with differential evolution (DE) technique. The differential evolution is a parallel direct search population based optimization method. From our simulations it is found that DE is able to optimize the quality measures of clusters of image datasets. To claim the superiority of DE based clustering we have compared the outcomes of DE with the classical K-means and popular Particle Swarm Optimization (PSO) algorithms for the same datasets. The comparisons results reveal the suitability of DE for image clustering in all image datasets.

Data Classification is a prime task in Data mining. Accurate and simple data classification task can help the clustering of large dataset appropriately. In this paper we have experimented and suggested a simple ANN based classification models called as Minimal ANN ( MANN) for different classification problems. The GA is used for optimally finding out the number of neurons in the single hidden layered model. Further, the model is trained with Back Propagation (BP) algorithm and GA (Genetic Algorithm) and classification accuracies are compared. It is revealed from the simulation that our suggested model can be a very good candidate for many applications as these are simple with good performances.

No abstract