HiCS: High Contrast Subspaces for Density-Based Outlier Ranking

By Fabian Keller, Emmanuel Müller, Klemens Böhm. IEEE International Conference on Data Engineering, 2012

377 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/7bd05d7130b3cb64d3f8c8481810e4a2e8b935f6

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Statistics and Probability

Citations in the corpus

Outlier Detection with Autoencoder Ensembles (2017, methodology): “For HiCS the number of Monte Carlo trials was set to 80, α = 0.1, and candidate cutoff was set to 40.”
Progress in Outlier Detection Techniques: A Survey (2019, methodology): “[85] proposed a high contrast subspace method (HiCS) to improve on evaluating and ranking of outliers where outlier scores are closely related.”
Contextual Outlier Interpretation (2017, methodology): “We follow the procedures in [Keller et al., 2012] and create two synthetic datasets with ground-truth abnormal attributes for Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence (IJCAI-18)”
Robust Subspace Clustering (2013, background): “Finally, in the data mining literature subspace clustering is sometimes used to describe a different—although related—problem; see [26, 28, 39].”
A survey on unsupervised outlier detection in high‐dimensional numerical data (2012, background): “The latest paper on the topic of subspace outlier detection is dedicated to identifying high contrast subspaces (HiCS) [112], using Monte Carlo sampling to obtain a notion of contrast of a subspace.”
Outlier Analysis (2013, methodology): “In the work proposed in [256], the outliers are found only in these high-contrast subspaces, and the corresponding scores are combined together.”
Hierarchical Density Estimates for Data Clustering, Visualization, and Outlier Detection (2015, methodology): “…and search outliers in subspaces of the data space [Aggarwal and Yu 2001; Kriegel et al. 2009b; Müller et al. 2008, 2010, 2011, 2012; Nguyen et al. 2011; Keller et al. 2012; Micenková et al. ACM Transactions on Knowledge Discovery from Data, Vol. 10, No. 1, Article 5, Publication date: July 2015.”
Deep Anomaly Detection with Deviation Networks (2019, background): “g irrelevant features or non-linear separable classes due to the curse of dimensionality and the deficiency in capturing the non-linear relations.Recentlyensemblemethods(e.g.,iForest[14]andmanyothers [9, 18]) showed some large improvement over these approaches by working on selected feature subspace”
Deep Learning for Anomaly Detection (2020, background): “. Performing anomaly detection in a reduced lower-dimensional space spanned by a small subset of original features or newly constructed features is a straightforward solution, e.g., in subspace-based [71, 78, 86, 124] and feature selection-based methods [12, 111, 113, 113]. However, identifying intr”
Outlier Detection for Text Data (2017, methodology): “The key methods, which are used frequently for outlier analysis include distance-based methods [13, 20], density-based methods [4], and subspace methods [2, 11, 16, 19, 15].”
Loda: Lightweight on-line detector of anomalies (2016, methodology): “Detectors with sub-space sampling were recently proposed (Keller et al. 2012; Nguyen et al. 2010; Muller et al. 2011) but neither the robustness against missing features or the explanation of the cause of a sample's anomalousness had been mentioned.”
Graph based anomaly detection and description: a survey (2014, positive): “2013), and subspace-based (Keller et al. 2012; Kriegel et al. 2012; Müller et al. 2010, 2012) techniques.”
Deep Autoencoding Gaussian Mixture Model for Unsupervised Anomaly Detection (2018, background): “(2011)), complex system management (Liu et al. (2008)), medical care (Keller et al.”
A comprehensive survey of anomaly detection techniques for high dimensional big data (2020, background): “Fabien and Kelloer [100] proposed to estimate the contrast of subspaces by enhancing the quality of traditional anomaly rankings by calculating the scores of high-contrast projections as evaluated on real data sets.”