Extreme Value Theory for Open Set Classification - GPD and GEV Classifiers
By Edoardo Vignotto, Sebastian Engelke. 2018
19 citations recorded by the source corpus.
Classification tasks usually assume that all possible classes are present during the training phase. This is restrictive if the algorithm is used over a long time and possibly encounters samples from unknown classes. The recently introduced extreme value machine, a classifier motivated by extreme value theory, addresses this problem and achieves competitive performance in specific cases. We show that this algorithm can fail when the geometries of known and unknown classes differ. To overcome this problem, we propose two new algorithms relying on approximations from extreme value theory. We show the effectiveness of our classifiers in simulations and on the LETTER and MNIST data sets.
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
- Jihong Park
- Medhi Bennis
- Sumudu Samarakoon
- Merouane Debbah
- Mingzhe Chen
- Walid Saad
- Songcan Chen
- Chuanxing Geng
Citations in the corpus
- Wireless Network Intelligence at the Edge (2018, background): “Using the principles of EVT, the tail distribution of the queue lengths exceeding a predefined threshold is characterized by a GPD Gd(·) with two parameters d = [σ, ξ] scale and shape, respectively.”
- Recent Advances in Open Set Recognition: A Survey (2018, methodology): “To address these limitations, Vignotto and Engelke [75] further presented the GPD and GEV classifiers relying on approximations from EVT.”