Cluster-based outlier detection

Abstract

Outlier detection has important applications in the field of data mining, such as fraud detection, customer behavior analysis, and intrusion detection. Outlier detection is the process of detecting the data objects which are grossly different from or inconsistent with the remaining set of data. Outliers are traditionally considered as single points; however, there is a key observation that many abnormal events have both temporal and spatial locality, which might form small clusters that also need to be deemed as outliers. In other words, not only a single point but also a small cluster can probably be an outlier. In this paper, we present a new definition for outliers: cluster-based outlier, which is meaningful and provides importance to the local data behavior, and how to detect outliers by the clustering algorithm LDBSCAN (Duan et al. in Inf. Syst. 32 (7): [978][979][980][981][982][983][984][985][986] 2007) which is capable of finding clusters and assigning LOF (Breunig et al. in

ZERO++: Harnessing the Power of Zero Appearances to Detect Anomalies in Large-Scale Data Sets

“…Examples are distance based methods (e.g., k NN, see Ramaswamy, Rastogi, & Shim, 2000; LDOF, see Zhang, Hutter, & Jin, 2009), density based methods (e.g., LOF, see Breunig et al, 2000), clustering based methods (e.g., CBOF, see Duan, Xu, Liu, & Lee, 2009), and isolation based methods (e.g., iForest, see Liu et al, 2012;MassAD, see Ting, Zhou, Liu, & Tan, 2013). These methods rely on the key characteristic of numeric data, i.e., the notion of ordering.…”

EMG Signal Decomposition Using Motor Unit Potential Train Validity

“…Using several MUP-shape features along with the MU firing pattern features used may also improve the performance of the SCC. The MUP-shape features can be extracted based on the methods presented for finding the outliers in a cluster [186][187][188] or even the methods developed for detecting super-imposed MUPs in a MUPT [189].…”

Unsupervised outlier detection in streaming data using weighted clustering

“…Outlier detection has been very interesting topic for research community [3]- [11]. Ramaswamy et al proposed a distance based outlier detection method in [3].…”

“…He et al in [5] presented new definition of outlier which they named as cluster-based local outlier, which provides importance to the local data behaviour. Duan et al in [11] proposed a cluster based outlier detection algorithm which can detect both single point outliers and cluster-based outliers.…”