Statistics@UPF

David Rossell will give a keynote talk at the BAYSM meeting organized by the junior section of International Society for Bayesian Analysis 

We invite applications for faculty appointments at the level of tenure-track Assistant Professor in the area of Statistics at the Department of Economics and Business to begin no later than September 2023.Applicants should hold a Ph.D. (or be near completion of their doctoral studies), and should have a demonstrated potential for research excellence in Statistics, Probability, Machine Learning or related fields. Applicants should provide a curriculum vitae, research statement, submit a copy 1-2 papers with their most significant work and 3 letters of recommendation.

The deadline for application is December 15, 2022. For more information, including how to apply, see here. For informal inquiries you may contact david.rossell@upf.edu

We are looking for two postdoctoral researchers to work on various topics related to the analysis of complex high-dimensional data. Topics of interest include but are not limited to new directions in high-dimensional statistics, statistical learning for non i.i.d. data, change-point detection, deep learning, and stochastic gradient descent. The positions are for one year, with a gross yearly salary of 35,000 eur. For more information see here.

We have two post-doctoral positions to work on a range of topics related to network analysis, ultra high-dimensional graphical models, frameworks linking networks and graphical models, multivariate structural learning and time series analysis. The positions are for 1 year, with a potential for extension to 18 months, with a gross yearly salary of 40,000 eur. For more information see here.

Jack Jewson obtained a Juan de la Cierva post-doctoral fellowship in Mathematics and Paul Rognon obtained an AGAUR FI grant to support his PhD studies. 

Alejandra Avalos-Pacheco, former PhD student of David Rossell, was awarded the Savage award (Applied Methodology) for the best PhD thesis by the International Society for Bayesian Analysis 

The phenomenon of benign overfitting is one of the key mysteries uncovered by deep learning methodology: deep neural networks seem to predict well, even with a perfect fit to noisy training data. Motivated by this phenomenon, Gábor and his co-authors consider when a perfect fit to training data in linear regression is compatible with accurate prediction. Their analysis shows that overparameterization is essential for benign overfitting in this setting: the number of directions in parameter space that are unimportant for prediction must significantly exceed the sample size. (link to the article)

The European Research Council (ERC) has awarded a Starting Grant to our colleague Gergely Neu. In his ERC project, Gergely will study reinforcement learning in artificial intelligence.

Gábor, Christian, Gergely, Geert, and Dani strike back with an amazing concert in Ovella Negra. Their cover of "War Pigs" will give you a glimpse of how it was.