Formal name: Prof.dr. P.D. Grunwald
Function: Scientific Staff Member
Email: Peter.Grunwald@cwi.nl
Telephone: +31(0)20 592 4115
Room: L235
Homepage: www.cwi.nl/~pdg
Research groups: Algorithms and Complexity
Research
Peter Grünwald's research interests lie where statistics, computer science and information theory meet: theories of learning from data. More specifically, he mainly works on (1) the minimum description length (MDL) principle and its variants - he is author of the comprehensive book The Minimum Description Length Principle, MIT Press, June 2007; (2) statistical algorithms for and analysis of the realistic situation in which `all models are wrong, yet some are useful' - from 2005 to 2010 he headed the NWO Vidi project called `learning when all models are wrong'. Recently, he has started working on (3) the risky but sometimes inevitable use of statistics in the courtroom. In his current Vici project 'safe statistics', (1), (2) and (3) all play a significant role.
Career
| 2013 - | Scientific staff member A&C - Algorithms and complexity |
| 2010 - 2013 | Scientific staff member PNA6 - Algorithms and Complexity |
| 2008 - | Full professor Leiden University |
| 1999 - 2009 | Scientific staff member INS4 - Quantum Computing and Advanced Systems Research |
Selected Awards and Honours
| 2010 | Vici Innovational Research Grant NWO |
| 2010 | Van Dantzig Prijs |
| 2005 | Vidi Innovational Research Grant NWO |
Selected Academic Activities
| 2010 | Program chair Catalina Island, CA, USA - International Conference on Uncertainty in Artificial Intelligence - [UAI] |
| 2008 | PhD advisor Rooij, S. de |
| 2008 - | Member steering committee PASCAL Network of Excellence - II |
| 2008 - 2010 | Member NWO, Den Haag - Reviewing Committee Veni Innovation Grant Proposals |
| 2007 | Guest editor Journal: EURASIP Journal on Bioinformatics and Systems Biology - Issue on Information-theoretic Methods for Bioinformatics |
| 2006 | PhD co-advisor Cilibrasi, R. |
| 2006 | Member program committee AAAI Spring Symposium on Distributed Plan and Schedule Management |
| 2006 - 2010 | PhD advisor Erven, T.A.L. van |
| 2004 - 2008 | Member steering committee PASCAL Network of Excellence |
Selected Publications
| R.D. Gill, P.D. Grünwald. An algorithmic and a geometric characterization of coarsening at random. Annals of Statistics 36, 2409–2422, 2008. |
| Tim van Erven, P.D. Grünwald, S. de Rooij. Catching up faster in Bayesian model selection and model averaging. 2008. |
| P.D. Grünwald. The minimum description length principle. , 1–570 MIT Press., 2007. |
| P.D. Grünwald, J. Langford. Suboptimal behavior of Bayes and MDL in classification under misspecification. Machine Learning 66, 119–149, 2007. |
| T. Roos, H. Wettig, P.D. Grünwald, P. Myllymaki, H. Tirri. On discriminative Bayesian network classifiers and logistic regression. Machine Learning 59, 267–296, 2005. |
