[agents] Open position: permanent research fellow in reinforcement learning at INRAE-TOULOUSE

[Meritxell Vinyals Salgado]

*INRAE **is recruiting**: **a **permanent position **for a **research
fellow** in **reinforcement learning** is open in our research unit MIAT,
located in **Toulouse, Occitanie (France)**.*

*To apply :*

*https://jobs.inrae.fr/en/open-competitions/open-competitions-research-scientists-job-profiles-crcn/cr-2025-mathnum-4
<https://jobs.inrae.fr/en/open-competitions/open-competitions-research-scientists-job-profiles-crcn/cr-2025-mathnum-4>*

*For any additional information,* please feel free to contact Régis
Sabbadin (regis.sabbadin at inrae.fr), scientific coordinator, or Nathalie
Vialaneix (nathalie.vialaneix at inrae.fr), unit director.

*Tentative recruitment timeline:*

   -

   *Application opening:* January 28, 2025
   -

   *Application deadline:* March 4, 2025
   -

   *Eligibility examination:* April 2025
   -

   *Final selection examination:* June 2–18, 2025
   -

   *Start date:* From September 1, 2025

*Your mission*: You will be positioned within the SCIDyn (Simulation,
Control and Inference of Agro-environmental and Biological Dynamics) team,
which consists of nine researchers and engineers, primarily from the fields
of computer science and statistics. One of the main research areas of the
SCIDyn team is reinforcement learning (RL), a rapidly expanding field in
designing decision models for agriculture, forestry, etc. Classical RL
approaches are applicable when a model of the system to be controlled is
available; however, when only observational data of the system are
accessible, inverse RL approaches are required. You will contribute to
strengthening the team's expertise in reinforcement learning in this area,
with a mission to develop innovative RL approaches and algorithms in
collaboration with SCIDyn researchers, within a context where experiments
are costly but data of varying types, quality, and complexity are available
from our agronomist and ecologist partners.

In particular, your research project will aim to develop and implement RL
algorithms using observed trajectories of controlled systems, where
observations come from various sources: direct experiments, simulation,
observations of other agents, experiments on different agroecosystems, etc.
You will explore one or more areas of RL (Inverse RL, Batch RL, imitation
RL, Deep RL, etc.) in alignment with the SCIDyn team’s finalized
objectives, with the support of team researchers skilled in single and
multi-agent RL.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cs.umbc.edu/pipermail/agents/attachments/20250204/f91ccc74/attachment-0001.htm>