Current facets (Pre-Master)

PhD defence of Tim Salimans on Thursday 23 May 2013

Promotie Tim Salimans op donderdag 23 mei 2013

On Thursday 23 May 2013 Tim Salimans will defend his PhD thesis entitled 'Essays in Likelihood-Based Computational Econometrics'. Supervisors are Professors  Richard Paap en Dennis Fok. Other members of the Doctoral Committee are professors John Geweke en Dick van Dijk (Erasmus School of Economics) and Siem Jan Koopman (VU Amsterdam).

Time and location

The PhD defence will take place in the Senate Hall of Erasmus University Rotterdam and will start at 13.30 hrs.

About the dissertation

Big data has been one of the business buzz-words of the last couple of years. However, techniques for analysing these data sets still lag behind this development, since sophisticated statistical models often require enormous computer resources to be applied to data sets of this size. In his dissertation, Tim Salimans develops new computational methods that make it easier to apply sophisticated statistical models to the data sets of today.

In  his dissertation, Tim also applies these new techniques to various statistical problems, including forecasting newspaper sales, locating dark matter in space, predicting the outcomes of chess matches, and recommending Xbox games.

Tim has shown the practical value of his research. He has won several statistical prediction competitions held by the San-Francisco based competitive predictive modelling platform Kaggle.com. This platform ranks him as one of the best data scientists in the world.  As a result, he has gained media coverage in several countries.

He advices various Dutch companies through his predictive analytics consulting firm, Algoritmica.

Tim has published in several journals, including the Journal of Econometrics and Bayesian Analysis.

About Tim Salimans

Tim Salimans (1985) obtained a BSc in Science from University College Utrecht in 2007 and received an MPhil in Economics from Tinbergen Institute in 2009. He continued his PhD research at the Tinbergen Institute / Erasmus School of Economics in Rotterdam. During this period, he spent three months at Microsoft Research Cambridge as part of his dissertation research. He currently works at predictive analytics consulting firm Algoritmica, which he co-founded in 2012.

Besides his academic research, Tim is also a regular competitor in statistical prediction competitions, held by Kaggle.com. In these competitions, organizations supply the competitors with a data set and a modelling problem, while holding back part of the data. The challenge is to build the best statistical model to predict the values of the data that is held back. The winner of these competitions receives a cash prize. Tim is the winner of several of these competitions, among others:

  • The Deloitte/FIDE Chess rating challenge, predicting the outcomes of chess matches based on past results. Sponsored by Deloitte and the world chess federation FIDE.
  • Observing Dark Worlds competition, discovering the location of dark matter halos in space, based on the distortions they cause in astronomical observations. Organized by the University of Edinburgh, sponsored by Winton capital.
  • Currently in 2nd place out of 1500+ teams for the Heritage Health Prize, a contest to predict hospitalizations with over 3 million dollars in total prize money.

Abstract of 'Essays in Likelihood-Based Computational Econometrics'

Econometrics relies on probabilistic models to describe and analyse how observed data relates to economic hypotheses and to provide a rigorous framework for reasoning under uncertainty. Statistical analysis of such models can be performed using likelihood-based inference methods such as Bayesian analysis and the method of maximum likelihood.

This dissertation deals with the computational challenges associated with these methods. It aims to solve these computational problems using Monte Carlo methods as well as deterministic approximations of the (marginal) likelihood. By developing efficient approximate inference algorithms, this work addresses various problems in economics for which likelihood-based econometric inference was previously difficult or infeasible.