Omiros Papaspiliopoulos

Omiros Papaspiliopoulos

Universitat Pompeu Fabra

Social & Behavioural Sciences

Previously to becoming ICREA Research Professor I had been Research Associate at Lancaster and Oxford University, Assistant Professor at Warwick University, and Professor at UPF. I am currently the director of the Masters in Data Science and the Data Science Center at Barcelona GSE. I have extensively published in the top journals in Statistics, have served as Associate Editor for several journals and as of January 2018 as Deputy Editor for Biometrika. I have delivered more than 100 invited talks & seminars, and given courses at ENSAE in Paris, the Berlin Mathematical School, the Department of Mathematics at University of Copenhagen, the Engineering Department at Osaka University, CEMFI. In 2010 I was awarded the Royal Statistical Society's Guy Medal in Bronze, which is arguably the highest distinction in Statistics in Europe. 

Research interests

I am involved with the whole spectrum of Data Science, from real practical problems to serious mathematics for developing and analyzing methods. My current research evolves along the following three non-orthogonal axes: Axis 1: Methodological work in the intersection of Statistics, Machine Learning and Applied Mathematics; Axis 2: Data Science within Social Sciences; Axis 3: Applied Machine Learning projects. Major subtheme in Axis 1  is scalable Bayesian computation. Bayesian models provide an excellent framework for synthesizing heterogenous data, learning and predicting with big but sparse data, and  principled regularization in high-dimensional models. Scalable Bayesian computation refers to computational methods for solving large scale Bayesian learning problems whose complexity scales favourably with the amount of data and parameters, ideally as a linear function of these two quantities. This is enourmously important in practice: it makes an approach realistic for large scale applications. Within Axis 2 I have been involved in a range of application that despite their apparent diversity they have important common components and strong links with Axis 1: predicting electoral outcomes in elections with emerging political parties, long-term forecasting of real estate prices at zip-code level;  understanding channels of central bank communication; surveilance and prediction of social unrest.  As founder and director of the Barcelona Graduate School of Economics Data Science Center (DSC), I have been involved in all the projects the DSC undertakes and fall within Axis 3. Recent collaborations include the Reuters Institute for the  Study of Journalism in Oxford, Accenture Health Analytics, Banco de España, ZDF (German public-service television broadcaster).

Selected publications

- Papaspiliopoulos O, Roberts GO & Zanella G 2020, 'Scalable inference for crossed random effects models', Biometrika, 2020, 107, Issue 1, Pages 25–40.

- Chopin N and Papaspiliopoulos O 2020, An introduction to Sequential Monte Carlo, Springer Series in Statistics. ISBN: 978- 3-030-47845-2.

- Papaspiliopoulos, Omiros 2020, 'High-Dimensional Probability: An Introduction with Applications in Data Science', Quantitative Finance, 20, 10, 1591 - 1594.

Selected research activities

Editorial activities and reviewing for major funding bodies:

  • Deputy Editor for Biometrika
  • Finnish Academy of Sciences: member of a six-people panel that decides on the whole funding program of the Academy for the scientific fields of Statistics and Applied Mathematics.
  • Associate Editor for SIAM Journal of Uncertainty Quantification

Conference organization:

Scientific Committee for Conference of the International Society of Non Parametric Statistics (ISNPS), Cyprus, June 2020

Academic Leadership:

  • Director of the BGSE Data Science Center
  • Director of the BGSE Master in Data Science
  • Director of the BGSE Summer School in Data Science
  • Director of the BGSE Winter School in Data Science