Next-generation optimal experimental design scalability and real world impact h/f

A propos d'Inria

Inria est l'institut national de recherche dédié aux sciences et technologies du numérique. Il emploie 2600 personnes. Ses 215 équipes-projets agiles, en général communes avec des partenaires académiques, impliquent plus de 3900 scientifiques pour relever les défis du numérique, souvent à l'interface d'autres disciplines. L'institut fait appel à de nombreux talents dans plus d'une quarantaine de métiers différents. 900 personnels d'appui à la recherche et à l'innovation contribuent à faire émerger et grandir des projets scientifiques ou entrepreneuriaux qui impactent le monde. Inria travaille avec de nombreuses entreprises et a accompagné la création de plus de 200 start-up. L'institut s'eorce ainsi de répondre aux enjeux de la transformation numérique de la science, de la société et de l'économie. Next-generation optimal experimental design : scalability, and real world impact
Le descriptif de l'offre ci-dessous est en Anglais
Type de contrat : CDD

Contrat renouvelable : Oui

Niveau de diplôme exigé : Thèse ou équivalent

Fonction : Ingénieur scientifique contractuel

Niveau d'expérience souhaité : De 3 à 5 ans

A propos du centre ou de la direction fonctionnelle

The Centre Inria de l'Université de Grenoble groups together almost 600 people in 22 research teams and 7 research support departments.

Staff is present on three campuses in Grenoble, in close collaboration with other research and higher education institutions (Université Grenoble Alpes, CNRS, CEA, INRAE), but also with key economic players in the area.

The Centre Inria de l'Université Grenoble Alpe is active in the fields of high-performance computing, verification and embedded systems, modeling of the environment at multiple levels, and data science and artificial intelligence. The center is a top-level scientific institute with an extensive network of international collaborations in Europe and the REST of the world.

Contexte et atouts du poste

Within the Statify team.

Mission confiée

Optimal Experimental Design (OED) is a powerful tool to reduce the cost of running a sequence of experiments. Myriad application areas have witnessed, on the one hand, an explosion in the volume of data that can BE acquired, and on the other, the use of increasingly complex and computationally intensive models to interpret these data. Yet large volumes of data do not by default carry a commensurate amount of information. Moreover, we inevitably face constraints : on the costs of experimentation or data acquisition, on data storage and communication, and on the computational effort of statistical inference in complex models. OED directly addresses the associated trade-offs--e.g., between experimentation, measurement, and/or processing costs and the quality of subsequent and decision making.This project aims at highlighting computational developments at the OED research frontier.

Principales activités

Methods for stochastic simulation, high-dimensional approximation or integration, and stochastic optimization are central to modern OED and to the scaling of OED to large parameter spaces and complex statistical models. Modern machine learning methodologies--from neural network surrogates, to deep reinforcement learning in sequential OED, to modern generative models and transport methods for simulation-based inference--also play a catalyzing role in such OED approaches.

Actvities will include the development in terms of methods and software of a number of these new OED methodologies, and showcase applications to real-world problems. As a first task, we will examin solutions based on the Expected Information Gain (EIG). In this latter case, design optimization corresponds to the maximization of some intractable expected contrast between prior and posterior distributions. Scaling this maximization to high dimensional and complex settings has been an issue due to OED inherent computational complexity. We aim at making this scaling possible by implementing state-of-the-art generative models adapted to OED starting from recent publications in the team.

Compétences

Expertise in statistics and machine learning and more specifically in generative models and their mathematical properties is necessary together with a high knowledge of Python/Jax.

Avantages
- Subsidized meals
- Partial reimbursement of public transport costs
- Leave : 7 weeks of annual leave + 10 extra days off due to RTT (statutory reduction in working hours) + possibility of exceptional leave (sick children, moving home, etc.)
- Possibility of teleworking (after 6 months of employment) and flexible organization of working hours
- Professional equipment available (videoconferencing, loan of computer equipment, etc.)
- Social, cultural and sports events and activities
- Access to vocational training
- Social security coverage