N. K. Nagwani

Sanjay Chakraborty, N. K. Nagwani

This paper describes the incremental behaviours of Density based clustering. It specially focuses on the Density Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm and its incremental approach.DBSCAN relies on a density based notion of clusters.It discovers clusters of arbitrary shapes in spatial databases with noise.In incremental approach, the DBSCAN algorithm is applied to a dynamic database where the data may be frequently updated. After insertions or deletions to the dynamic database, the clustering discovered by DBSCAN has to be updated. And we measure the new cluster by directly compute the new data entering into the existing clusters instead of rerunning the algorithm.It finally discovers new updated clusters and outliers as well.Thus it describes at what percent of delta change in the original database the actual and incremental DBSCAN algorithms behave like same.DBSCAN is widely used in those situations where large multidimensional databases are maintained such as Data Warehouse.

Sanjay Chakraborty, N. K. Nagwani, Lopamudra Dey

Incremental K-means and DBSCAN are two very important and popular clustering techniques for today's large dynamic databases (Data warehouses, WWW and so on) where data are changed at random fashion. The performance of the incremental K-means and the incremental DBSCAN are different with each other based on their time analysis characteristics. Both algorithms are efficient compare to their existing algorithms with respect to time, cost and effort. In this paper, the performance evaluation of incremental DBSCAN clustering algorithm is implemented and most importantly it is compared with the performance of incremental K-means clustering algorithm and it also explains the characteristics of these two algorithms based on the changes of the data in the database. This paper also explains some logical differences between these two most popular clustering algorithms. This paper uses an air pollution database as original database on which the experiment is performed.

Sanjay Chakraborty, N. K. Nagwani

The incremental K-means clustering algorithm has already been proposed and analysed in paper [Chakraborty and Nagwani, 2011]. It is a very innovative approach which is applicable in periodically incremental environment and dealing with a bulk of updates. In this paper the performance evaluation is done for this incremental K-means clustering algorithm using air pollution database. This paper also describes the comparison on the performance evaluations between existing K-means clustering and incremental K-means clustering using that particular database. It also evaluates that the particular point of change in the database upto which incremental K-means clustering performs much better than the existing K-means clustering. That particular point of change in the database is known as "Threshold value" or "% delta change in the database". This paper also defines the basic methodology for the incremental K-means clustering algorithm.