Network snomaly detection based on semi-supervised clustering
SMO'07 Proceedings of the 7th WSEAS International Conference on Simulation, Modelling and Optimization
Flexible Grid-Based Clustering
PKDD 2007 Proceedings of the 11th European conference on Principles and Practice of Knowledge Discovery in Databases
| Hi-index | 0.00 |
Few of the current existing methods for unsupervised learning (clustering) algorithms consider clustering the data points in a low-dimensional subspace in real time. In this paper, we present a grid based clustering algorithm (GCA) with time complexity (O(n)). Unlike previous clustering algorithm, GCA pays more attention to the running time of the algorithm. GCA achieves low running time by (i) determining the number of the clusters according to the point density of the grid cell and (ii) computing the distances between the centers of the clusters and the grid cells, not the data points. In order to make GCA more efficient, principal component analysis(PCA) is introduced to transform the data points from high dimension to low dimension. Finally, we analyze the performance of GCA and show that it outperforms most of the current state-of-the-art methods in terms of efficiency. In particular, it outperforms k-means algorithm by several orders in the running time.