Analysis of Clustering Algorithm for Outlier Detection in Data Stream

Outlier detection is an important data mining task, aiming at the discovery of elements that show significant diversion from the expected behavior. Data stream mining has poses different challenges for outlier detection like concept drift, huge size and evolutionary data from data streams. Clustering techniques for data stream which helps to create a similar group of data are used to cluster the similar data items in data streams and also used to detect the outliers from data stream, so they are called as cluster based outlier detection. Which provides advantages like less memory requirement, less time consumption and it results exact outliers. In data streams if an object does not obey the behavior of normal data object is called as outlier. We proposed a new framework for outlier detection in data streams, which is combination of Neighbour based outlier detection approach and clustering based approach for outlier detection in data streams which provides better output in terms of true outliers from data streams.