Szczegóły publikacji
Opis bibliograficzny
Axiomatic kernels on graphs for support vector machines / Marcin ORCHEL, Johan A. K. Suykens // W: Artificial neural networks and machine learning - ICANN 2019 : workshop and special sessions : 28th International Conference on Artificial Neural Networks : Munich, Germany, September 17–19, 2019 : proceedings / eds. Igor V. Tetko, [et al.]. — Cham : Springer Nature Switzerland, cop. 2019. — (Lecture Notes in Computer Science ; ISSN 0302-9743 ; 11731). — ISBN: 978-3-030-30492-8; e-ISBN: 978-3-030-30493-5. — S. 685–700. — Bibliogr. s. 699–700, Abstr. — Publikacja dostępna online od: 2019-09-09. — M. Orchel - pierwsza afiliacja: ESAT-STADIUS, KU Leuven, Belgium
Autorzy (2)
- AGHOrchel Marcin
- Suykens Johan A. K.
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 124593 |
|---|---|
| Data dodania do BaDAP | 2019-09-26 |
| Tekst źródłowy | URL |
| DOI | 10.1007/978-3-030-30493-5_62 |
| Rok publikacji | 2019 |
| Typ publikacji | materiały konferencyjne (aut.) |
| Otwarty dostęp |  |
| Wydawca | Springer |
| Konferencja | International Conference on Artificial Neural Networks 2019 |
| Czasopismo/seria | Lecture Notes in Computer Science |
Abstract
We solve the problem of classification on graphs by generating a similarity matrix from a graph with virtual edges created using predefined rules. The rules are defined based on axioms for similarity spaces. Virtual edges are generated by solving the problem of computing paths with maximal fixed length. We perform experiments by using the similarity matrix as a kernel matrix in support vector machines (SVM). We consider two versions of SVM: for inductive and transductive learning. The experiments show that virtual edges reduce the number of support vectors. When comparing to kernels on graphs, the SVM method with virtual edges is faster while preserving similar generalization performance.
