Research interests: Sequential and Markov Chain Monte Carlo methods for Bayesian Statistics. I am especially interested in the methodological side of the field: finding new algorithms to tackle computational problems arising in Bayesian Statistics, for instance in high-dimensional, non-linear dynamical systems, or multi-modal static problems. I also try to contribute to the theory supporting the methods, and to more computational aspects such as parallel computing.
I am currently looking for a postdoctoral fellow. The position would start in December 2016, with a two-year contract.
The position is focused on improving the sequential Monte Carlo methodology for inference in non-linear dynamical systems and Bayesian statistics. Emphasis will be put on gradient-based and transport-based ideas to improve parameter inference and predictions. Applications in population growth models and disease spread modelling will be part of the project. Related ideas can be found in my recent "Couplings of Particle Filters" technical report.
Applicants must have received, or expect to complete, a doctorate in Statistics or related discipline within the last 5 years. Selection criteria will include demonstrated research ability in computational statistics and excellent verbal and written communication skills.
Applications should be sent by email before September 15, 2016, with a CV, a reference letter and a 1-page statement of research.
Email: pjacob at fas.harvard.edu.
Phone: (617) 496-9259
Office: Science Center 712, 1 Oxford Street, Cambridge, MA 02138