Indicative Support Vector Clustering with its Application on Anomaly Detection

In many learning scenarios, supervised learning is hardly applicable due to the unavailability of a complete se t of data labels, while unsupervised model overlooks valuable u ser feedback in an interactive system setting. In this paper, a n ovel semi-supervised support vector clustering algorithm is pr esented, where a small number of user indicated labels are available as supervised information. We apply the clustering algorit hm in the anomaly detection area, and show that the given labels significantly improve the recognition of anomalies. Moreov er, the partially labeled data proliferates the information witho ut extra computation but strengthening the robustness to anomalies .