A New Approach for Detecting Outliers in Data Streams

To continuous flow of data and it is a process of extracting knowledge structure from continuous, rapid data records and it can be considered as a subfield of data mining. Data Stream can be classified into two types they are offline and online streams. Online data stream used in an amount of real world appliances, including network traffic monitoring, intrusion detection, credit card and fraud detection and offline data stream are used in reports based on web log streams. Data size is extremely huge and potentially infinite and it’s not possible to store all the data, so it leads to a mining challenge where shortage of limitations occurs in hardware and software. Data mining techniques are newly proposed for data streams they are highly helpful to mine the data are data stream clustering, data stream classification, frequent pattern technique, sliding window techniques and so on. For outlier detection data stream clustering technique is highly desirable one. The main objective of this research work is to perform the clustering process in data streams and detecting the outliers in data streams. Two types of clustering algorithms namely FUZZY C-MEANS and CLARANS are used for finding the outliers in data streams. The two performance factors such as clustering accuracy and outlier detection accuracy are used for analysis. By analyzing the experimental results, it is observed that the CLARANS clustering algorithm performance is more accurate