Parameter-Free Outlier Scoring Algorithm Using the Acute Angle Order Difference Distance

Pumruckthum, Pollaton; Boonsiri, Somjai; Sinapiromsaran, Krung

doi:10.1007/978-3-030-19861-9_4

Cited by 2 publications

(5 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Various outlier scoring algorithms have been proposed in the literature [22][23][24] based on different metrics. An improved parameter-free outlier scoring algorithm based on the distance metric and angle is proposed [25] that avoids the weakness from [22][23][24]. However, using only the distance metric will give low scores to outliers forming a small clump.…”

Section: The Parameter-free Outlier Detectionmentioning

confidence: 99%

“…Acute angle order difference distance outlier factor (AOF) [24] algorithm measures outlier scores similar to the unblocked different distance of OOF but adds a condition that the acute angle of reference data point 𝑝 𝑟 to a pair of cover data point 𝑝 𝑖 and measured data point 𝑝 𝑗 . The AOF algorithm does not use the order different distance method for all data points.…”

Section: Acute Angle Order Difference Distance Outlier Factor (Aof)mentioning

confidence: 99%

“…However, the grid search technique cannot be used for unsupervised learning. To avoid using user-defined parameters, the methodology of measuring outlier scores by the order difference distance is proposed to be AOF [24], OOF [22], and WOF [23]. However, they fail to measure outlier scores correctly because these algorithms are based on the first nearest neighbor (𝑘 = 1 in the kNN algorithm) which is distance-based.…”

Section: The Motivation Of the Mof Algorithmmentioning

confidence: 99%

“…However, they fail to measure outlier scores correctly because these algorithms are based on the first nearest neighbor (𝑘 = 1 in the kNN algorithm) which is distance-based. They cannot detect small clusters of outliers and also fail to detect the edge of normal regions as outliers, according to Figure 4.2 from [24]. The radius of the red circle is the outlier score.…”

Section: The Motivation Of the Mof Algorithmmentioning

confidence: 99%

“…The MOF algorithm is compared to six other algorithms: LOF [41], OOF [22], AOF [24], FastABOD [25], and kNN, all of which are discussed in detail in Chapter 2. LOF, kNN, and FastABOD have a single parameter (𝑘-nearest neighbors), whereas OOF and AOF are parameter-free.…”

Section: Competitive Algorithmsmentioning

confidence: 99%

See 4 more Smart Citations

Parameter-free outlier scoring using mass ratio variance for static and streaming data

Changsakul

View full text Add to dashboard Cite

Outlier detection is a significant problem that has been studied in a variety of research and real-world applications. However, little research has been conducted on unsupervised parameter-free outlier scoring. This thesis proposes Mass ratio variance-based Outlier Factor, or MOF, which is unsupervised parameter-free outlier scoring for static data. This algorithm calculates outlier scores based on the variance of mass ratio. The data points with high outlier scores are associated with outliers while the data points with low outlier scores are associated with normal data points. This thesis also proposes Streaming Mass ratio variance-based Outlier Factor or SMOF. This algorithm calculates outlier scores based on MOF and the non-overlapping sliding window model which keeps the dense data points by weighted random sampling making highly efficient storage. Extensive experiments�have�been conducted to evaluate the performance of MOF and SMOF using synthesized and real-world data sets. The experimental results show that they have better accuracy than the state-of-the-art outlier detection techniques.

show abstract