Robust Model-Based Learning via Spatial-EM Algorithm

Yu, Kai Ming; Dang, Xin; Bart, Henry L.; Chen, Yixin

doi:10.1109/tkde.2014.2373355

Cited by 16 publications

(6 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The most straightforward outlier detection method, modelbased method, is to create a model for all samples, and then predict outliers as those having large deviations from the established profiles. For example, the Gaussian mixture model (GMM) [11] fits the whole dataset to a mixed Gaussian distribution and evaluates the parameters through the Expectation-Maximization [29] or a deep estimation network [12]. However, GMM needs to predetermine the appropriate cluster type and number, which are crucial and extremely difficult.…”

Section: Classic Outlier Detection Methodsmentioning

confidence: 99%

Generative Adversarial Active Learning for Unsupervised Outlier Detection

Liu

Zhou

et al. 2019

IEEE Trans. Knowl. Data Eng.

200

129

View full text Add to dashboard Cite

Outlier detection is an important topic in machine learning and has been used in a wide range of applications. In this paper, we approach outlier detection as a binary-classification issue by sampling potential outliers from a uniform reference distribution. However, due to the sparsity of data in high-dimensional space, a limited number of potential outliers may fail to provide sufficient information to assist the classifier in describing a boundary that can separate outliers from normal data effectively. To address this, we propose a novel Single-Objective Generative Adversarial Active Learning (SO-GAAL) method for outlier detection, which can directly generate informative potential outliers based on the mini-max game between a generator and a discriminator. Moreover, to prevent the generator from falling into the mode collapsing problem, the stop node of training should be determined when SO-GAAL is able to provide sufficient information. But without any prior information, it is extremely difficult for SO-GAAL. Therefore, we expand the network structure of SO-GAAL from a single generator to multiple generators with different objectives (MO-GAAL), which can generate a reasonable reference distribution for the whole dataset. We empirically compare the proposed approach with several state-of-the-art outlier detection methods on both synthetic and real-world datasets. The results show that MO-GAAL outperforms its competitors in the majority of cases, especially for datasets with various cluster types or high irrelevant variable ratio. The experiment codes are available at: https://github.com/leibinghe/GAAL-based-outlier-detection

show abstract

Section: Classic Outlier Detection Methodsmentioning

confidence: 99%

Generative Adversarial Active Learning for Unsupervised Outlier Detection

Liu

Zhou

et al. 2019

IEEE Trans. Knowl. Data Eng.

200

129

View full text Add to dashboard Cite

show abstract

“…Thus, researchers attempt to create a model representing the normal data and identify the data that deviates significantly from the established normal behaviors as anomalies. Among various methods, densitybased [29], [30], [31], [32], [33], clustering-based [34], [35], distance-based methods [9], [10] and ensemble detectors [36], [37], [38], [39] are the most popular methods. Density-based methods usually estimate the density function of the data, and then identify the anomalies as those having large deviations from the density peaks.…”

Section: Traditional Anomaly Detection Methodsmentioning

confidence: 99%

“…Density-based methods usually estimate the density function of the data, and then identify the anomalies as those having large deviations from the density peaks. For example, given a training set, the Gaussian mixture model (GMM) [29] fits a given number of Gaussian distributions to the data via the Expectation-Maximization (EM) [30] algorithm. The data points with the smallest likelihood are identified as anomalies.…”

Section: Traditional Anomaly Detection Methodsmentioning

confidence: 99%

Supervised Anomaly Detection via Conditional Generative Adversarial Network and Ensemble Active Learning

Chen¹,

Duan

Li³

et al. 2021

Preprint

View full text Add to dashboard Cite

Anomaly detection has wide applications in machine intelligence but is still a difficult unsolved problem. Major challenges include the rarity of labeled anomalies and it is a class highly imbalanced problem. Traditional unsupervised anomaly detectors are suboptimal while supervised models can easily make biased predictions towards normal data. In this paper, we present a new supervised anomaly detector through introducing the novel Ensemble Active Learning Generative Adversarial Network (EAL-GAN). EAL-GAN is a conditional GAN having a unique one generator vs. multiple discriminators architecture where anomaly detection is implemented by an auxiliary classifier of the discriminator. In addition to using the conditional GAN to generate class balanced supplementary training data, an innovative ensemble learning loss function ensuring each discriminator makes up for the deficiencies of the others is designed to overcome the class imbalanced problem, and an active learning algorithm is introduced to significantly reduce the cost of labeling real-world data. We present extensive experimental results to demonstrate that the new anomaly detector consistently outperforms a variety of SOTA methods by significant margins. The codes are available on Github.

show abstract

“…Yu et al [8] have presented Spatial -EM algorithm for finite mixture learning procedures. Median based location and rank-based scatter estimators that replaces sample mean and covariance in each M step to enhance the stability and robustness of the algorithm.…”

Section: Distribution Based Clusteringmentioning

confidence: 99%

An Enhanced Equilin Multi-Clustering Algorithm for Constructing Numeric Clusters

A¹,

S²

2016

Int. Jour. Info. Sci. Comp.

View full text Add to dashboard Cite

Clustering techniques are often applied in data analytics for interpreting the similarities within data objects over large datasets. Despite the existence of many clustering algorithm in the literature such as connectivity, centroid, distribution, density etc, the factor that constitutes a cluster are different from one another. However, the success of clustering depends upon the maximization of intra-cluster similarity and inter-cluster dissimilarity. The significant implication of clustering algorithms in many real-world applications emerges the proposal of newer algorithms. As a consequence, in this paper, a novel effort is made to generate clusters from a different aspect of grouping data objects using multidimensional geographical linear equation called Equilin Clustering. The technique incorporates the standard linear equation and the method of percentage split for clustering numerical data. The results show that the performance of Equilin Clustering yields better cluster results with reduced complexity over time and number of iterations.

show abstract

Robust Model-Based Learning via Spatial-EM Algorithm

Cited by 16 publications

References 49 publications

Generative Adversarial Active Learning for Unsupervised Outlier Detection

Generative Adversarial Active Learning for Unsupervised Outlier Detection

Supervised Anomaly Detection via Conditional Generative Adversarial Network and Ensemble Active Learning

An Enhanced Equilin Multi-Clustering Algorithm for Constructing Numeric Clusters

Contact Info

Product

Resources

About