One-class classification

By David Martinus Johannes Tax. 2001

630 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

A systematic study of the class imbalance problem in convolutional neural networks (2017, methodology): “One-class versions were developed for decision trees [80], random forests [81], k-NN and other types of classifiers [82].”
McPAD: A multiple classifier system for accurate payload-based anomaly detection (2009, background): “, unlabeled outliers), and allows us to obtain a more precise description of the target class [31].”
Measuring discrimination in algorithmic decision making (2017, background): “Ideas, how to apply those for more than two groups, can be borrowed from multi-class classiﬁcation (Bishop 2006), the multi-label classiﬁcation (Tsoumakas and Katakis 2007), and one-class classiﬁcation (Tax 2001) literature.”
Data Mining for Imbalanced Datasets: An Overview (2005, background): “The one-class learners provide an interesting alternative to the traditional discriminative approach, where in the classifier is learned on the target class alone (Japkowicz, 2001b, Juszczak and Duin, 2003, Raskutti and Kowalczyk, 2004, Tax, 2001 ).”
Identification and classification of microaneurysms for early detection of diabetic retinopathy (2013, methodology): “GMM and SVM classifiers are based on the work presented in [26,31], respectively.”
Content-based visual search learned from social media (2012, background): “One might consider bypassing the negative labeling problem by one-class learning, which creates classifiers using positive examples only [116].”
Toward Supervised Anomaly Detection (2014, methodology): “Moreover, dual optimization as proposed in the work of Tax (2001) cannot be considered an sound alternative due to possible duality gaps.”
Design of robust classifiers for adversarial environments (2011, unknown): “like anomaly-based intrusion detection [17].”
Latent-Class Hough Forests for 3D Object Detection and Pose Estimation (2014, background): “However, as observed by Tax, these methods suffer from the added issue of specifying the multi-dimensional margin of such a boundary to balance between false positives and negatives and incorrect assignment can significantly affect performance [34].”
Editorial: special issue on learning from imbalanced data sets (2004, background): “Mainly, two classes of learners were previously studied in the context of the recognition-based one-class approach— SVMs [50][49] and autoencoders [26][38]—and were found to be competitive [38].”
Using an Ensemble of One-Class SVM Classifiers to Harden Payload-based Anomaly Detection Systems (2006, background): “the outlier class may be motivated by the fact that it is too difficult or expensive to obtain a significant number of training patterns of this class [23].”
A Survey of Recent Trends in One Class Classification (2009, methodology): “Tax [2] considers a Polynomial and a Gaussian kernel and found that the Gaussian kernel works better for most data sets.”
One-and-a-Half-Class Multiple Classifier Systems for Secure Learning Against Evasion Attacks at Test Time (2015, background): “However, one-class classifiers may exhibit a significantly lower accuracy in the absence of attack (in particular, in high-dimensional feature spaces), as they do not exploit any information on the (available) malicious training data [10].”