Real Time Anomaly Detection Using Ensembles

Finding anomalous behavior of user in networks is crucial, in analysis of such behavior to identify the real user is very complicated. Classification is one technique for identifying the anomalous behavior. The anomaly detection rate can be improved by ensemble the different classifiers. Empirically, ensembles tend to yield better results when there is a significant diversity among the models. The available models all are on synthetic data. This paper analyzes the ensemble model to identify the anomaly in real time with improved accuracy.