Odabk: An Effective Approach to Detecting Outlier in Data Stream

Abstract

Currently, data mining in data stream becomes a very popular research field. One of the central tasks in mining data streams is that of identifying outliers which can lead to discovering unexpected and interesting knowledge, which is critical important. To effectively mine outliers in data stream, ODABK, an algorithm for outlier detection in data stream is presented. It is based on KNN and significantly enhanced by means of other data structures and its optimized logical operations. Finally, the paper reports experiments on a real-world census data which show that ODABK is more effective in detection rate and execution times.

Challenges and Issues in DATA Stream: A Review

“…This paper [15] focus on an importing issue to mine frequent item sets from huge data sets for data stream so for this purpose design an algorithm named AICFI, this algorithm very better than other algorithm (RLS-U) and used the idle time O(N) of system in a very better way and produce throughput larger than other algorithm comparably and does not require out of core (summary structure) ,and support large data and low level support. Another study by [16] focuses on an importing issue of outlier detection in data stream so for this purpose introduce an algorithm named ODABK this algorithm based on classification method that is on K-NN that is on statistics based and instance base store all training samples so its detection rate of outlier is high as well as take less time compare to other algorithm (LOADED). Another study by [17] focuses on an importing issue of keeping information of music which is in continuous streams that is to keep latest information of music which is necessary in frequent temporal fashion of pattern through online system so for this purpose design an algorithm named FTP-MQS so it is one pass algorithm and its working consists of three phases 1st take to initialize the window, 2nd sliding of window, 3rd one and last phase to generate frequent pattern temporally so it is a best algorithm in term of efficiency to mine frequent pattern temporary for music stream.…”

Association rules based algorithm for identifying outlier transactions in data stream

“…The characteristic of data stream is that the data arrive in timely order and continuously [11,33]. Since data streams applications have received lots of attention [13], a variety of data analysis researches are presented to find valuable information or to predict future trends. For example, Manku et al proposed framework of frequent pattern mining in data stream [22].…”