Junior research scientist in process statistics

Back to campaign's jobs listing

INRAE presentation

INRAE, the French National Research Institute for Agriculture, Food, and Environment, is a public research organization bringing together 12,000 employees across 272 units in 18 centers across France. As the world’s leading institute specializing in agriculture, food, and the environment, INRAE plays a key role in supporting the necessary transitions to address global challenges.

Faced with population growth, food security challenges, climate change, resource depletion, and biodiversity loss, INRAE is committed to developing scientific solutions and supporting the evolution of agricultural, food, and environmental practices.

INRAE is recruiting researchers by open competition and offering permanent position.

Work environment, missions and activities

You will work within the MISTEA unit based in Montpellier. MISTEA develops original methods in mathematics, computer science, and statistics applied to the environment and agronomy. Our main objective is to contribute to a better understanding of and provide decision support on issues related to agriculture and the management and protection of natural resources. MISTEA members have expertise in algorithms and stochastic processes, statistics, filtering and observers, optimization and segmentation, and their work addresses key challenges in digital livestock farming, bioprocesses, and phenotyping in sensor data processing. You will be associated with the “dynamical systems” research team and will work closely with the “probability-statistics” research team.
To enable the transition to more sustainable agricultural practices, the development of new sensor technologies is essential. This is why these technologies are increasingly being deployed in experimental stations, fields, and livestock farms. These diverse and heterogeneous data sources enable both a better understanding of agricultural systems and the deployment of low-cost technologies for farmers. However, the specific characteristics of this data and the underlying stochastic processes prevent AI methods from producing accurate predictions. Combining learning methods with mechanistic models is becoming an increasingly important challenge that requires expert knowledge.
You will contribute to this rapidly emerging field. Your work will focus on researching, studying, and implementing state-of-the-art numerical methods for the inference and calibration of random dynamic models for data from sensors in livestock farming and agriculture. You will contribute to the design and implementation of stochastic algorithms on these topics using appropriate learning methods (intensive computing, modeling, simulation, etc.) in order to improve understanding
of agricultural systems and provide digital tools for decision-making. You will participate in ongoing projects and develop your own to mobilize expert knowledge, increase the quality and robustness of predictions, and access unobservable quantities. You will participate in disseminating your methods to scientific partners and other stakeholders.
You will be involved in:
- developing methods for analyzing data from sensors coupled with mechanistic models, and studying their properties,
- analyzing random dynamic models that include Markov processes, diffusion processes, hidden Markov processes, or deterministic models,
- establish an inferential framework for estimating latent variables, such as breaks or mixture structures, using Bayesian, non-parametric, or high-dimensional statistical methods,
- interact with researchers from different fields, including statistics and artificial intelligence, dynamic systems modeling, and agronomy,
- disseminate your work in the form of publications in journals in your discipline and in more specialized disciplines, as well as in the form of packages (R, Python, Julia, etc.).
You will be expected to be able to implement some of the methods mentioned above and apply them to issues studied by the unit.Il sera attendu de pouvoir mettre en place quelques-unes des méthodes citées ci-dessus pour les appliquer à des problématiques étudiées par l’unité.

Training and skills

Competition open to candidates with a PhD (or equivalent). A PhD in applied mathematics is highly recommended. The successful candidate will be expected to have a good knowledge of
recent methods in process statistics: adaptive methods for estimating parameters (functional
or otherwise) of stochastic processes, inference of hidden Markov models or latent variable models, algorithms for
detecting breaks in diffusive or jump processes, variational approaches, etc. Knowledge of
Bayesian statistics and associated particle algorithms would be appreciated. An interest in
biological or environmental issues and experience in real-world applications would be
appreciated. The ability to communicate and collaborate with specialists in applied disciplines is expected.
Fluency in English is desirable, as is long-term international experience: successful candidates who do not yet have such experience will be required to spend time abroad at the end of the internship year.

INRAE's life quality

By joining our teams, you benefit from:

- 30 days of annual leave + 15 days "Reduction of Working Time" (for a full time);
- parenting support: CESU childcare, leisure services;
- skills development systems: training, career advise;
- social support: advice and listening, social assistance and loans;
- holiday and leisure services: holiday vouchers, accommodation at preferential rates;
- sports and cultural activities;
- collective catering.

For international scientists: please visit your guide to facilitate your arrival and stay at INRAE. 

How to apply

1. I download the applicant guide Guide for applicants 2026 pdf - 1.41 MB
2. I write down the profile number CR26-MathNum-1
3. I apply GO