Support Vector Data Description
Citations
9,627 citations
1,899 citations
1,266 citations
1,007 citations
876 citations
Cites methods from "Support Vector Data Description"
...For the explanation below, two of the most common 1SVM algorithms are chosen, a hypersphere-based 1SVM (known as Support Vector Data Description (SVDD)) by Tax and Duin [8], and a Planebased 1SVM (PSVM) by Scholkopf et al. [31], see Fig....
[...]
...The parameters of SVM based methods are selected via a grid-search, width ν ð0 1Þ, and σ ð1 1Þ for SVDD [8], and γ ð2 (15);2 (13);....
[...]
...For the explanation below, two of the most common 1SVM algorithms are chosen, a hypersphere-based 1SVM (known as Support Vector Data Description (SVDD)) by Tax and Duin [8], and a Planebased 1SVM (PSVM) by Scholkopf et al....
[...]
...This is extended from the hypersphere-based one-class SVM approach proposed by Tax and Duin [8]....
[...]
...Further, Tax and Duin have shown that the hyperplane-based one-class SVM becomes a special case of the (equivalent) hypersphere-based scheme when used with a radial basis kernel....
[...]
References
40,147 citations
"Support Vector Data Description" refers methods in this paper
...This is identical to the approach which is used in Schölkopf, Burges, and Vapnik (1995) to estimate the VC-dimension of a classifier (which is bounded by the diameter of the smallest sphere enclosing the data)....
[...]
26,531 citations
"Support Vector Data Description" refers background or methods in this paper
...Several kernel functions have been proposed for the Support Vector Classifier (Vapnik, 1998; Smola, Schölkopf, & Müller, 1998)....
[...]
...In contrast to the Support Vector Classifier, the Support Vector Data Description using a polynomial kernel suffers from the large influence of the norms of the object vectors, but it shows promising results for the Gaussian kernel....
[...]
...For that the notion of essential support vectors has to be introduced (Vapnik, 1998)....
[...]
...Vapnik argued that in order to solve a problem, one should not try to solve a more general problem as an intermediate step ( Vapnik, 1998 )....
[...]
...The classifiers are Gaussian-density based linear classifier (called Bayes), Parzen classifier and the Support Vector Classifier with polynomial kernel, degree 3....
[...]
19,056 citations
13,033 citations
"Support Vector Data Description" refers background or methods in this paper
...Neural networks, for instance, can be trained to estimate posterior probabilities (Richard & Lippmann, 1991; Bishop, 1995; Ripley, 1996) and tend to give high confidence outputs for objects which are remote from the training set....
[...]
...The third method is a Mixture of Gaussians, optimized using EM (Bishop, 1995)....
[...]
...By applying Leave-One-Out estimation (Vapnik, 1998; Bishop, 1995), it can be shown that the number of support vectors is an indication of the expected error made on the target set....
[...]
...In classification or regression problems a more advanced Bayesian approach can be used for detecting outliers (Bishop, 1995; MacKay, 1992; Roberts & Penny, 1996)....
[...]
...Keywords: outlier detection, novelty detection, one-class classification, support vector classifier, support vector data description...
[...]