Technique for Density Based Clustering Using Neural Network in Data Mining

The clustering is the technique in which similar and dissimilar type of data is clustered in different clusters for batter analysis of the input data. The algorithm of DBSCAN is applied in which EPS is calculated which will be the central point and from the central point Euclidean distance is calculated to define similarity and dissimilarity of the input data. In the existing algorithm EPS is calculated dynamically but Euclidian distance statically which reduce accuracy of clustering. In this work, back propagation algorithm is been applied which calculate Euclidian distance dynamically and simulation study is conducted which shows that proposed improvement increase accuracy of clustering and reduce execution time. Keywords— Clustering, DBSCAN, Back-propagation, Accuracy, Execution time