Uniform Object Generation for Optimizing One-class Classifiers

By David Martinus Johannes Tax, Robert P. W. Duin. Journal of machine learning research, 2002

354 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/0bb0c6dfeb8f838470f27cefd0ba42192ecaf90a

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

PEBL: positive example based learning for Web page classification using SVM (2002, background): “One-class SVMs have been recently developed, which distinguish one class of data from the rest of the feature space given only positive data set 19, 14.”
Support vector machines (2008, unknown): “Alternatively (Tax and Duin, 2001), the target class can be modeled by fitting a hypersphere with minimal radius around it.”
Latent-Class Hough Forests for 3D Object Detection and Pose Estimation (2014, background): “Conversely, other approaches attempt to artificially generate a negative, outlier, dataset [19, 32, 77], however the generalization ability can depend heavily on the procedure used for data generation and carry the risk of adding unintentional bias into the negative set.”
PEBL: Web page classification without negative examples (2004, background): “from the rest of the feature space given only positive data sets [ 24 ], [22].”
One-class classification: taxonomy of study and review of techniques (2013, unknown): “In practical implementations, the method of Scho¨lkopf et al . and the SVDD method of Tax and Duin (2001b) operate comparably and both perform best when the Gaussian kernel is used.”
Novelty detection: a review - part 2: : neural network based approaches (2003, background): “Tax and Duin (2001) suggest creating outliers uniformly in and around the target class.”
Support vector machines for predicting distribution of Sudden Oak Death in California (2005, methodology): “By using kernels, one-class SVMs are able to represent various data distribution shapes in feature space (e.g. banana shapes, sphere shapes, or even very irregular shapes; Tax and Duin, 2002).”
A Survey of Recent Trends in One Class Classification (2009, methodology): “OSVMs have been successfully applied in a wide variety of application domains such as Handwritten Digit Recognition [1][2][19], Information Retrieval [20], Face Recognition Applications [28][37], Medical Analysis [38], Bioinformatics [39][40], Spam Detection [41], AnomalyDetection [42][43], Machine ”
Social Media Text Classification under Negative Covariate Shift (2015, methodology): “In order to balance between model over-fitting and under-fitting, Tax and Duin (2001) proposed a method that tries to use artificially generated outliers to optimize the model parameters.”
A systematic study of the class imbalance problem in convolutional neural networks (2017, background): “particle swarm optimization [74, 75]) or outliers detection strategy [76, 77].”