Comparative Analysis of Clustering Algorithms for Outlier Detection in Data Streams

Nowadays, data mining has become one of the most popular research areas in the field of computer science, because data mining techniques are used for extracting the hidden knowledge from the large databases. In data mining, most of the work is emphasized over knowledge discovery and data stream mining is becoming an active research area in this domain. A data stream is a similar to river, it means continuous and massive sequence of data elements are in and out generated at a rapid rate and the analysis of data stream has been recently attracted attention over in data mining research community. When the amount of data is very huge, it leads to a numerous computational and mining challenges due to shortage of hardware and software limitations. Data mining techniques are newly proposed for data streams they are highly helpful to mine are data stream clustering, data stream classification, frequent pattern technique, sliding window techniques and so on. For outlier detection data stream clustering algorithm is highly needed. This main objective of this research work is to perform the clustering process in data streams and detecting the outliers in data streams. In this research work, two clustering algorithms namely BIRCH with CLARANS and CURE with CLARANS are used for finding the outliers in data streams. Different types, sizes of data sets and 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 CURE with CLARANS clustering algorithm performance is more accurate than the BIRCH with CLARANS.