Biography

PhD in Applied Mathematics, focusing on Artificial Intelligence, Machine Learning, numerical methods for Probability and Statistics and on applications to Finance. I obtained my PhD in 2023 at LPSM Sorbonne Université (formerly UPMC - University Pierre and Marie Curie) under the direction of Gilles Pagès.

Since April 2024 I have been working at Smart News in the Vertical Machine Learning team, focusing on news understanding through artificial intelligence for recommendation.

Interests
  • Machine Learning
  • Stochastic Optimization
  • Numerical Probability
  • Monte Carlo methods
  • Mathematical Finance
  • Stochastic control and Reinforcement Learning
Education

  • PhD in Applied Mathematics, 2020-2023

    Sorbonne Université, LPSM

  • Master 2 Probability and Random Models, option Applied Probability, 2019

    Sorbonne Université, LPSM

  • Diploma of Ecole Normale Supérieure in Mathematics and Applications, 2016-2020

    Ecole Normale Supérieure de la rue d'Ulm

Publications & pre-Publications

Pierre Bras (2023). Adaptive Gradient Langevin Algorithms for Stochastic Optimization and Bayesian Inference. PhD thesis.

PDF Cite Slides

Pierre Bras, Gilles Pagès (2023). Policy Gradient Optimal Correlation Search for Variance Reduction in Monte Carlo simulation and Maximum Optimal Transport. In arXiv e-prints.

PDF Cite arXiv GitHub

Pierre Bras, Gilles Pagès (2022). Langevin algorithms for Markovian Neural Networks and Deep Stochastic control. In International Joint Conference on Neural Networks IJCNN23.

PDF Cite IJCNN23 arXiv GitHub Slides Poster

Pierre Bras (2022). A note on $L^1$-Convergence of the Empiric Minimizer for unbounded functions with fast growth. In arXiv e-prints.

PDF Cite arXiv

See all publications

Recent & Upcoming Talks

Séminaire de Mathématiques Appliquées du Collège de France - Pierre-Louis Lions
Algorithmes adaptatifs de gradient-Langevin pour l’optimisation stochastique et l’inférence Bayésienne
Dec 8, 2023 11:15 AM — 12:30 PM Collège de France
14th International Conference on Monte Carlo Methods and Applications
Convergence of Langevin-Simulated Annealing algorithms with multiplicative noise
Jun 28, 2023 12:00 PM Sorbonne Université Campus Pierre et Marie Curie
International Neural Network Society Workshop on Deep Learning Innovations and Applications INNS DLIA 2023
Langevin Algorithms for Very Deep Neural Networks with application to Image Classification
Jun 23, 2023 Gold Coast Convention and Exhibition Centre, Australia
See all events

Code & Jupyter Notebooks

relocor: REinforcement Learning Optimal CORrelation search
Implementation of the package relocor: REinforcement Learning Optimal CORrelation search, a stochastic control and reinforcement-learning based method for variance reduction in Monte Carlo simulation of stochastic differential equations.
relocor: REinforcement Learning Optimal CORrelation search
Langevin algorithms for Markovian Neural Networks and Deep Stochastic control
In this repository we give the implementation of Langevin and Layer Langevin optimizers as instances of the TensorFlow tf.keras.optimizers.Optimizer base class and we compare Langevin and non-Langevin optimizers for the training of various stochastic control problems.
Langevin algorithms for Markovian Neural Networks and Deep Stochastic control
Langevin algorithms for very deep Neural Networks with applications to image classification
In this repository we give the implementation of Langevin and Layer Langevin optimizers as instances of the TensorFlow tf.keras.optimizers.Optimizer base class and we compare Langevin and non-Langevin optimizers for the training of various problems for image classification.
Langevin algorithms for very deep Neural Networks with applications to image classification
See all

Contact