Analyzing and Optimizing ANT-Clustering Algorithm by Using Numerical Methods for Efficient Data Mining

Clustering analysis is an important function of data mining. There are various clustering methods in DataMining. Based on these methods various clustering algorithms are developed. Ant-clustering algorithm isone of such approaches that perform cluster analysis based on “Swarm Intelligence’. Existing antclusteringalgorithm uses two user defined parameters to calculate the picking-up probability and droppingprobability those are used to form the cluster. But, use of user defined parameters may lead to form aninaccurate cluster. It is difficult to anticipate about the value of the user defined parameters in advance toform the cluster because of the diversified characteristics of the dataset. In this paper, we have analyzedthe existing ant-clustering algorithm and then numerical analysis method of linear equation is proposedbased on the characteristics of the dataset that does not need any user defined parameters to form theclusters. Results of numerical experiments on synthetic datasets demonstrate the effectiveness of theproposed method.