Curriculum vitae

Stoehr Julien

Associate Professor
CEREMADE

stoehrping@ceremade.dauphinepong.fr
Phone : 4967
Office : B640
Personal URL

Biography

Julien Stoehr has been a lecturer at Université Paris-Dauphine since 2017. After graduating from the École Normale Supérieure Rennes (former École Normale Supérieure de Cachan - Antenne de Bretagne) and receiving an external degree (agrégation) in Mathematics, he got a PhD in statistics and specialized in the field of computational statistics. His research focuses on the design and implementation of statistical methodologies in a Bayesian setting where the statistical model is complex, e.g., likelihood with no closed form or generative models that are time consuming to simulate from, with a particular interest for the Monte Carlo methods (Markov chain Monte Carlo, Hamiltonian Monte Carlo and approximate Bayesian methods (ABC)).

Latest publications

Articles

Robert C., Stoehr J. (2024), Simulating signed mixtures, Statistics and Computing, vol. 35, n°9

Clarte G., Robert C., Ryder R., Stoehr J. (2021), Component-wise approximate Bayesian computation via Gibbs-like steps, Biometrika, vol. 108, n°3, p. 591–607

Stoehr J., Benson A., Friel N. (2018), Noisy Hamiltonian Monte Carlo for Doubly Intractable Distributions, Journal of Computational and Graphical Statistics, vol. 28, n°1, p. 220-232

Stoehr J., Marin J-M., Pudlo P. (2016), Hidden Gibbs random fields model selection using Block Likelihood Information Criterion, Stat, vol. 5, n°1, p. 158-172

Stoehr J., Pudlo P., Cucala L. (2015), Adaptive ABC model choice and geometric summary statistics for hidden Gibbs random fields, Statistics and Computing, vol. 25, n°1, p. 129-141

Communications avec actes

Stoehr J., Friel N. (2015), Calibration of conditional composite likelihood for Bayesian inference on Gibbs random fields, in , Volume 38: Artificial Intelligence and Statistics, 9-12 May 2015, San Diego, California, USA, IEEE - Institute of Electrical and Electronics Engineers, 921-929 p.

Communications sans actes

Batardière B., Chiquet J., Kwon J., Stoehr J. (2024), A gradient approximation with importance sampling for dimension reduction in natural exponential families, 55èmes journées de Statistiques de la SFdS, Bordeaux, France

Prépublications / Cahiers de recherche

Stoehr J., Robin S. (2024), Composite likelihood inference for the Poisson log-normal model, Paris, Cahier de recherche CEREMADE, Université Paris Dauphine-PSL, 32 p.

Batardière B., Chiquet J., Kwon J., Stoehr J. (2024), Importance sampling-based gradient method for dimension reduction in Poisson log-normal model, Paris, Cahier de recherche CEREMADE, Université Paris Dauphine-PSL, 44 p.

Stoehr J. (2019), A review on statistical inference methods for discrete Markov random fields, Paris, Cahier de recherche CEREMADE, Université Paris Dauphine-PSL, 31 p.

Wu C., Stoehr J., Robert C. (2019), Faster Hamiltonian Monte Carlo by Learning Leapfrog Scale, Paris, Cahier de recherche CEREMADE, Université Paris Dauphine-PSL, 18 p.

Clarte G., Ryder R., Robert C., Stoehr J. (2019), Component-wise approximate Bayesian computation via Gibbs-like steps, Paris, Cahier de recherche CEREMADE, Université Paris Dauphine-PSL, 30 p.

Back to the list