Výsledky vyhledávání - "Michaël Allouche"

A refined Weissman estimator for extreme quantiles

Autor: Michaël Allouche, Jonathan El Methni, Stéphane Girard

Publikováno v: EcoSta 2022-5th International Conference on Econometrics and Statistics
EcoSta 2022-5th International Conference on Econometrics and Statistics, Jun 2022, Kyoto, Japan
Extremes
Extremes, In press
Compstat 2022-24th International Conference on Computational Statistics
Compstat 2022-24th International Conference on Computational Statistics, Aug 2022, Bologna, Italy

International audience; Weissman extrapolation methodology for estimating extreme quantiles from heavy-tailed distributions is based on two estimators: an order statistic to estimate an intermediate quantile and an estimator of the tail-index. The co

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::ca612f4a8882324462792f478b255eb3
https://hal.science/hal-03688584

Zobrazit plný text záznamu

EV-GAN: Simulation of extreme events with ReLU neural networks

Autor: Michaël Allouche, Stéphane Girard, Emmanuel Gobet

Publikováno v: EcoSta 2022-5th International Conference on Econometrics and Statistics
EcoSta 2022-5th International Conference on Econometrics and Statistics, Jun 2022, Kyoto, Japan
Journal of Machine Learning Research
Journal of Machine Learning Research, 2022, 23 (150), pp.1--39
HAL

International audience; 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 parametriz

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::db82266ec3b87490afe65c6bbe46c5db
https://hal.archives-ouvertes.fr/hal-03250663v3

Zobrazit plný text záznamu

On the approximation of extreme quantiles with ReLU neural networks

Autor: Michaël Allouche, Stéphane Girard, Emmanuel Gobet

Publikováno v: EVA 2021-12th International Conference on Extreme Value Analysis
EVA 2021-12th International Conference on Extreme Value Analysis, Jun 2021, Edinburgh / Virtual, United Kingdom
HAL

International audience; Feedforward neural networks based on rectified linear units (ReLU) cannot efficiently approximate quantile functions which are not bounded in the Fréchet maximum domain of attraction. We thus propose a new parametrization for

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::76c48acc160085a6e9f7be32e22f0742
https://hal.science/hal-03301431

Zobrazit plný text záznamu

On the approximation of extreme quantiles with neural networks

Autor: Michaël Allouche, Stéphane Girard, Emmanuel Gobet

Publikováno v: SFdS 2021-52èmes Journées de Statistique de la Société Française de Statistique
SFdS 2021-52èmes Journées de Statistique de la Société Française de Statistique, Jun 2021, Nice, France. pp.1-5
HAL

National audience; In this study, we propose a new parametrization for the generator of a Generative adversarial network (GAN) adapted to data from heavy-tailed distributions. We provide an analysis of the uniform error between an extreme quantile an

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::3434d1c3409fcb3b21c787e481344125
https://inria.hal.science/hal-03268702

Zobrazit plný text záznamu

Generative model for fBm with deep ReLU neural networks

Autor: Michaël Allouche, Stéphane Girard, Emmanuel Gobet

Publikováno v: Bernoulli-IMS 2021-10th World Congress in Probability and Statistics
Bernoulli-IMS 2021-10th World Congress in Probability and Statistics, Jul 2021, Seoul / Virtual, South Korea
HAL

International audience; Over the last few years, a new paradigm of generative models based on neural networks have shown impressive results to simulate – with high fidelity – objects in high-dimension, while being fast in the simulation phase. In

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::818379394174993084dc59d4f09e876c
https://hal.archives-ouvertes.fr/hal-03237854

Zobrazit plný text záznamu

Akademický článek

Prevalence and association of placental lesions with obstetrical features and outcome: data from French prospective studyAJOG Global Reports at a Glance

Autor: Jeremy Boujenah, Jonathan Cohen, Michael Allouche, Marianne Ziol, Amélie Benbara, Marion Fermaut, Olivier Fain, Lionel Carbillon, Arsène Mekinian

Publikováno v: AJOG Global Reports, Vol 4, Iss 3, Pp 100374- (2024)

ABSTRACT: Purpose: Since the Consensus Statement diffused by the Amsterdam Placental Workshop Group, knowledge of the meaning of placental vascular malperfusion has become essential in the unavoidable analysis of obstetrical history in a patient foll

Externí odkaz: https://doaj.org/article/5e4d8ad60de5451e8e38825aaeccb331

Zobrazit plný text záznamu

Vyhledávací nástroje:

Upřesnit hledání