Clustering Algorithm of Data Mining to Detect Network Statistics

The augmentation growth in network data is causing a contemplative problem of detecting the useful information from the network. In reality the security technologies are not the final solution to prevent from security breaches. The important role of data mining is to detect the patterns of attacks in the network. Since, with a lot of technological preferment, the distinct numbers of attacks are proliferating day by day. Now, cryptography is not commensurate to save the supersensitive information. On the ground to interdict the network attacks their skeletons are identified from KDD 1999 dataset scenting the sequestered data of user's interest in a very less execution time using the data mining tool WEKA. In this proposed study, we are engaging the distinct clustering algorithms such as Farthest-first, Make-Density, Simple K-Means and Filtered Cluster, procuring the statistics of number of attacks found and what are the percentages of attacks on different network layer protocols.