En poursuivant votre navigation sur ce site, vous acceptez le dépôt de cookies dans votre navigateur. (En savoir plus)

PhD in Machine Learning by stochastic approaches: application to water clusters (H/F)

This offer is available in the following languages:
Français - Anglais

Date Limite Candidature : vendredi 21 mai 2021

Assurez-vous que votre profil candidat soit correctement renseigné avant de postuler. Les informations de votre profil complètent celles associées à chaque candidature. Afin d’augmenter votre visibilité sur notre Portail Emploi et ainsi permettre aux recruteurs de consulter votre profil candidat, vous avez la possibilité de déposer votre CV dans notre CVThèque en un clic !

Faites connaître cette offre !

General information

Reference : UMR7590-MICCAS-002
Workplace : PARIS 05
Date of publication : Friday, April 30, 2021
Scientific Responsible name : Michele Casula
Type of Contract : PhD Student contract / Thesis offer
Contract Period : 36 months
Start date of the thesis : 1 October 2021
Proportion of work : Full time
Remuneration : 2 135,00 € gross monthly

Description of the thesis topic

Thanks to recent progress, machine learning (ML) methods have shown extraordinary predictive capabilities in a great variety of domains, including chemistry, physics, and quantitative finance. In this project, we address the specific target of combining ML methods with stochastic approaches. We will build new ML schemes which are not only robust against noise-dominated training datasets, but could also exploit randomness to accelerate the learning process, by exploiting the stochastic occurrence of events obtained from unbiased sampling.
We will lay down the mathematical bases of this stochastic semi-supervised framework, and we will apply it to simulate water clusters, a notoriously challenging system in quantum chemistry, where appropriate training sets will be generated by both classical and quantum Monte Carlo methods.
A fully consistent description of water that builds from the fundamental interactions between hydrogen, oxygen and their surrounding electrons is still lacking. It is necessary to include nuclear quantum effects, whose motion must develop on the top of accurate potential energy surfaces (PES), provided by the quantum solution of the electronic problem at given nuclear coordinates. However, the computational cost of such an approach makes it difficult to study fully quantum bulk water. In this PhD project, we will adapt semi-supervised ML techniques to generate new interatomic potentials, with the aim of machine learning water clusters from quantum2 (quantum dynamics + quantum Monte Carlo) -based simulations[1-3]. Merging together these two worlds for the first time will allow us to draw an unprecedented picture of water, which is historically among the main scientific interests of the laboratory[4,5].
[1] F. Mouhat et al. 2017 Journal of Chemical Theory and Computation 13, 2400
[2] K. Nakano et al. 2020, The Journal of Chemical Physics 152, 204121
[3] K. Nakano et al. 2021, Phys. Rev. B 103, L121110
[4] A.M. Saitta et al. 2012, PRL 108, 207801
[5] S. Pipolo et al. 2017, PRL 119, 245701

Work Context

The thesis will be carried out in the TQM group, a young and dynamic theory group in the multidisciplinary IMPMC institute, located in the lively Pierre and Marie Curie campus of Sorbonne University. The TQM team gathers people who are world leaders in the development of new numerical methods from first principles and who made several publications in high-impact scientific journals[6,7].
[6] B. W. Lebert et al. 2019, PNAS 116, 20280
[7] D. Santos-Cottin et al. 2016, Nature Communications 7, 11258

We talk about it on Twitter!