Performance enhancement of DBSCAN density based clustering algorithm in data mining

Clustering is a form of learning by observations. It is an unsupervised learning method and does not require training data set to generate a model. Clustering can lead to the discovery of previously unknown groups within the data. It is a common method of data mining in which similar and dissimilar type of data would be clustered into different clusters for better analysis of the data. In this paper the DBSCAN algorithm has been applied to compute the EPS value and Euclidian distance on the basis of similarity or dissimilarity of the input data. Also back propagation algorithm is applied to calculate Euclidian distance dynamically and simulation study is conducted that shows improvement to increase accuracy and reduce execution time.