Workshop on Statistical and Machine
Learning Techniques in Computer
Intrusion Detection
Fuzzy Gravitational Clustering for Intrusion Detection
Jonatan Gomez and Dipankar Dasgupta
Abstract:
This paper proposes a soft computing technique for solving intrusion detection
problems that combines an unsupervised-robust clustering technique based
on gravitational concepts [1] with a fuzzy cluster analysis. In this approach,
a set of clusters, for the normal samples, is generated using the gravitational
clustering algorithm, then fuzzy membership functions are assigned to
these clusters (of normal samples) in order to indicate the degree of
mormalcy. Experiments with real data sets are performed in order to show
the applicability of the proposed approach and to compare with other works
reported in the literature.