Résumé
Feedforward neural networks based on Rectified linear units (ReLU) cannot efficiently approximate quantile functions which are not bounded, especially in the case of heavy-tailed distributions. We thus propose a new parametrization for the generator of a Generative adversarial network (GAN) adapted to this framework, basing on extreme-value theory. An analysis of the uniform error between the extreme quantile and its GAN approximation is provided: We establish that the rate of convergence of the error is mainly driven by the second-order parameter of the data distribution. The above results are illustrated on simulated data and real financial data. It appears that our approach outperforms the classical GAN in a wide range of situations including high-dimensional and dependent data.
| langue originale | Anglais |
|---|---|
| journal | Journal of Machine Learning Research |
| Volume | 23 |
| état | Publié - 1 mai 2022 |
Empreinte digitale
Examiner les sujets de recherche de « EV-GAN: Simulation of extreme events with ReLU neural networks ». Ensemble, ils forment une empreinte digitale unique.Contient cette citation
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver