A New Outlier Detection Algorithm Based on Manifold Learning

摘要：

Detecting outliers in a large set of data objects is a major data mining task aiming at finding different mechanisms responsible for different groups of objects in a data set. All existing approaches, however, are based on an assessment of distances (sometimes indirectly by assuming certain distributions) in the full-dimensional Euclidean data space. In high-dimensional data, these approaches are bound to deteriorate due to the notorious curse of dimensionality In this paper, we propose a novel approach named MLOD (Manifold Learning -Based Outlier Detection), This way, the effects of the curse of dimensionality are alleviated compared to purely distance-based approaches. A main advantage of our new approach is that our method does not rely on any parameter selection in.uencing the quality of the achieved ranking. Empirical studies conducted on both real and synthetic data sets show that signi.cant improvements in detection rate and false alarm rate are achieved using the proposed framework.

关键词： Manifold Learning LIE algorithm Outlier detection

作者: Zhigang Tang Jun Yang Bingru Yang

作者单位: School of Information Engineering, University of Science and Technology Beijing, Beijing 100083,Chin School of Information Engineering, University of Science and Technology Beijing, Beijing 100083,Chin

会议类型: 国际会议

会议名称: The 22nd China Control and Decision Conference(2010年中国控制与决策会议)

会议地点: 徐州

会议语种:英文

页码: 452-457

在线出版日期: 2010-05-26（万方平台首次上网日期，不代表论文的发表时间）