Marie Skłodowska-Curie Actions

Université Claude Bernard Lyon 1- Hosting offer for an MSCA-IF post-doc candidate in Gaussian Process Modelling in Actuarial Science

    24/07/2019 18:00 - Europe/Athens
    H2020 / Marie Skłodowska-Curie Actions
    France, Villeurbanne
    Université Claude Bernard Lyon 1
    Actuarial Science and Finance Institute
    Laboratory of Actuarial Science and Finance

The Laboratoire de Sciences Actuarielle et Financière (LSAF) is a leading actuarial research lab in Europe and undertakes inter-disciplinary research on risks in insurance and finance. The research topics of the laboratory are always evolving to include new risks (human longevity improvements, natural hazards...), recent accounting standards (IFRS), new prudential regulation systems (Basel 4, Solvency 2) as well as new risk management practices (Enterprise Risk Management) or new business models with Big Data and Analytics in insurance. Current team research projects concern in particular the impact of modeling and data analytics on the management of insurance and financial firms, environmental risks, longevity risk management and public health, long term investments and ESG, as well as behaviours and risks.

Strongly associated to ISFA graduate school of actuarial studies, LSAF gathers 30 permanent researchers, 4 post-docs and 18 PhD students. LSAF is part of University Claude Bernard Lyon 1 (UCBL), a multidisciplinary university in the primary fields of science and health. With over 2630 professors and assistant professors and 39,000 students, leading to more than 4500 internationally published articles and 40 patents per year. The University is involved in more than 80 European Union projects.

The LSAF offers to host a MSCA Individual Fellowship candidate ("Experienced Researcher" according to the Marie Sklodowska Curie categories, typically a post-doc), submitting an application to the next MSCA-IF call for proposals (deadline september 2019), interested to work on the following research topic:

Gaussian Process Modelling is sometimes used in Actuarial Sciences in order to model the output of large Economic Scenario Generator, to find Economic Capital Requirements (quantiles), or to optimise asset allocations or other quantity of interest (see [1], [2], [3] for example). In such models, the output is modelled by a Gaussian Process, with a given covariance structure or kernel. Selecting a kernel is essential to correctly represent the sample paths properties of the output (e.g. continuity, differentiability). Once selected, (hyper)parameter estimation of that kernel is a keypoint. In particular, it is crucial to account for uncertainty in estimation: it can be shown that, starting from few design points, optimization procedures that are simply based on pointwise kernel parameter estimation converge more slowly that full Bayesian ones [4].

The present project will develop new ideas for kernel parameter estimation, with a focus on uncertainty, using for instance specific cross-validation tools. We are looking for a candidate with a strong background in machine learning and actuarial sciences. Her/his PhD defence must have taken place before Sept. 10th, 2019.


The fellowship could last for 12 to 36 months, depending on the type of Individual Fellowship.


[1] Rullière, D., Faleh, A., Planchet, F., Youssef, W. (2013), Exploring or reducing noise ?, a global optimization algorithm in the presence of noise. Structural and Multidisciplinary Optimization, vol. 47, 6, pp. 921-936.

[2] Cousin, A., Maatouk, H., Rullière, D. (2016), Kriging of financial term-structures, European Journal of Operational Research, vol. 255, issue 2, pp. 631-648

[3] Jones, D. R., Schonlau, M., & Welch, W. J. (1998). Efficient global optimization of expensive black-box functions. Journal of Global optimization, 13(4), 455-492.

[4] Emmanuel Vazquez. Sequential search strategies based on kriging . Computation [stat.CO]. Université Paris-Sud, 2015. 〈tel-01266334〉


The successful Marie-Curie Post-doctoral fellow will be supervised by Didier Rullière and Olivier Roustant.

Didier Rullière is associate professor, University Lyon 1, actuary, and member of the LSAF laboratory. His research interests include dependence modelling, spatial statistics and risk measures. https://www.researchgate.net/profile/Didier_Rulliere

Olivier Roustant is Professor in Applied Mathematics at Mines Saint-Etienne, a school of Mines-Telecom Institute, member of the LIMOS research laboratory and associate member of the LSAF Laboratory. His main research domain are probability and statistics, statistical/machine learning, artificial intelligence. https://olivier-roustant.fr/ .


Application process to LSAF:

Interested candidates are invited to contact us exclusively by email at postdoc-1904@isfa.fr. Please do not write directly to the related researchers.

Make sure that you include the reference "Post-doc 1904" in the title of your email. Please attach a CV, a motivation letter, your MSc marks, as well as a 1-page research proposal.

NB: Candidates will receive the support of the LSAF supervisors, as well as a professional grant application company, to prepare and submit their application with the LSAF as a host laboratory, to the next MSCA-IF call for proposals.


The responsibility for the hosting offers published on this website, including the hosting description, lies entirely with the publishing institutions. The application is handled uniquely by the employer, who is also fully responsible for the recruitment and selection processes.