Abstract: Time series data streams are common due to the increasing usage of wireless sensor networks. Such data are often accompanied with uncertainty due to the limitations of data collection equipment. Outlier detection on uncertain static data is a challenging research problem in data mining. Moreover, the continuous arrival of data makes it more challenging. In this paper propose continuous outlier detection is a special class of steam data mining. Typically, stream data mining algorithms assume that each object is inspected at most once. However, in continuous outlier detection need to be capable of reporting, at each time point, the outliers among all the objects in the current sliding window. The propose a sliding window approach of outlier detection, which makes use of the results obtained from the previous state set to efficiently detect outliers in the current state set. These methods are verified using both real data and synthetic data. The results show that they are able to reduce the required storage and running time.

Keywords: Intrusion Detection Systems, Sliding window, MCOD, Event window.