CLIQUE clustering approach to detect denial-of-service attacks

We propose a grid based technique to mine the KDD Cup ´99 data. We propose a novel idea of using mixed clustering technique called clustering in quest (CLIQUE) (R. Agrawal et al., 1998) in experiments with KDD Cup ´99 data to detect attacks efficiently. Novelty lies in the fact that CLIQUE was never used on network traffic data. The results produced by CLIQUE when evaluated on synthetic data sets improved as the dimensionality of the data increased. Based on these results we assumed that CLIQUE can handle large database of high dimensional network traffic data efficiently. CLIQUE clustering technique is a combination of grid-based clustering and density-based clustering (R. Agrawal et al., 1998).