PAREEKSHA - A Machine Learning Approach for Intrusion and Anomaly Detection

Membership functions help us to identify and know the similarity between two elements such as vectors or sequences. The objective of this paper is to suggest a membership function and apply this membership function for learning the nature of dataset. In the initial learning process, the element vectors obtained are grouped to obtain clusters. The grouping is carried using incremental clustering technique. The initial knowledge thus build is later validated using the extended membership function so that any wrongly classified elements are placed properly. We name the approach as PAREEKSHA. The membership function is obtained by extending the basic Gaussian membership function and is inspired by approaches such as CLAPP, G-SPAMINE, and GARUDA in the recent research literature.