Missions

The purpose of the project is to design fully automated and data-driven procedures for pointwise and/or credibility interval estimation of epidemiological indicators, e.g., for the reproduction number R(t) of Covid19. Elaborating on a recent epidemiological model, both variational estimators and Monte Carlo samplers have been designed and implemented during the pandemic to estimate the reproduction number of the Covid19. The major bottleneck to their systematic use and generalization to other epidemics is that they require fine-tuning of hyperparameters, which until now has been done manually in conjunction with experts, inducing a prohibitive complexity. Automated data-driven selection procedures will enable to gain objectivity and capacity to handle large amount of data from a wide range of epidemics.
The first challenge consist in refining previous models to better account for both the epidemiological mechanisms and the possibly low quality of data reported during an epidemic. Mutliplicative models will be considered and connection with Kullback-Leibler Non-Negative Matrix Factorization will be explored. The second challenge is to leverage the derived statistical models to design automated data-driven procedures for the estimation of epidemiological indicators. To that aim, both Stein-based bilevel optimization, empirical Bayesian and unsupervised deep learning approaches will be considered.

Activities

The recruited postdoc researcher will tackle both implementation challenges and theoretical questions related to statistical modeling, prior design in the Bayesian framework, convex and non convex optimization, stochastic optimization. He/she is expected to develop commented, easy to handle codes to make available the proposed methodologies to nonspecialists. He/she will work in contact with epidemiologists and will be provided real epidemiological data. An interest in interdisciplinary research will be highly appreciated.

Skills

Prospective applicants are expected to hold a PhD in signal processing, statistics or a related discipline, excellent programming skills (e.g., in Python or Matlab), and good communication skills in English, both written and oral.

Work Context

The recruited candidate will be hired by the Centre National de la Recherche Scientifique (CNRS) in the framework of ANR grant OptiMoCSI holded jointly by LP-IXXI in Lyon, IMT in Toulouse and LS2N in Nantes. CNRS is the largest state-funded French research institution, employing researchers in all fields from exact sciences to humanities. He/she will integrate the Laboratoire des Sciences du Numérique de Nantes (LS2N), in the Signal, Image and Sound (SIMS) team (https://www.ls2n.fr/equipe/sims/) and work on the campus of Centrale Nantes, a top-level engineering school. “Regularly quoted in newspapers as being one of the nicest cities in France, Nantes is also renowned for being a rich, lively and innovative city. Its economic clout makes Nantes France's 3rd largest industrial city and 2nd most successful city in terms of employment growth." (https://metropole.nantes.fr/nouveaux-arrivants)
The contract is one year and it will be possible to renew it once.

Constraints and risks

N/A