An efficient technique for outlier detection in data mining.
- 1, ,
The Outlier detection problem has important applications in the field of fraud detection, network robustness analysis and intrusion detection. Most of such uses are high dimensional domains in which the data can contain hundreds of dimensions. Many recent algorithms employed the concept of proximity in order to find the outliers based on their relationships to the rest of data. But in high dimensional space, the data is sparse and the notation of proximity fails to retain its meaning. In fact the sparsity of high dimensional data, the notation of finding meaningfull outliers becomes substantially more complex and non-obvious. This article is an attempt to enhance outlier detection and analysis, which is an interesting data mining task. This outlier mining has a plethora of applications in fraud detection and for finding abnormal responses in various fields. In computer based outlier mining there are many methods and techniques. This work is an attempt to highlight the merits and demerits on the application of these methods. This would enable to find a suitable technique for perfect outlier detection. The main objective of this present article is to find out the best approach by comparative analysis.
Keywords: data mining, dataset, fraud detection, outlier detection
Citation: *, ( 2007), An efficient technique for outlier detection in data mining.. Scientific Transactions in Environment and Technovation Journal(STET), 1(1): 23-28
Received: 2015-06-17 15:53:32; Accepted: 2015-06-17 15:56:55;
*Correspondence: unknow user, unknow user