Niels Richard Hansen
Professor in Computational Statistics
Department of Mathematical Sciences
Universitetsparken 5
DK-2100 Copenhagen Ø
+45 - 35 32 07 83
In contemporary statistical applications we face model uncertainty as well as parameter uncertainty. Model search techniques are not just a possibility but in many applications a necessity. The key challenge is to develop methods that give an honest assessment of uncertainty. Underestimation of uncertainty leads to overconfident decisions, underestimation of risk, inflation of false discovery rates etc., which can have highly unfortunate consequences in any single application.

My main interest is in understanding how model uncertainty can be assessed and translated into usable methodology. I work on methods for data analysis that are both applicable and scalable to large-scale data analysis while at the same time amenable to mathematical analysis. The methods find applications to predictive as well as causal modeling.

I am currently leading the research project Causal inference with incomplete data, funded by VILLUM FONDEN.

Research subjects include: multivariate and high-dimensional dynamic models; model selection and penalized estimation methods; causal dynamic models; machine learning; R. I am, moreover, a cofounder of the The Dynamical Systems Interdisciplinary Network.
Past research interests include applied and computational probability theory, Markov chains and biological sequence analysis (2000-2009). From 2010 I have focused more on statistics, in particular aspects related to computational statistics, R software development and model selection questions.
Master's in Mathematics (2000), PhD in Statistics (2004), University of Copenhagen.
Assistant and Associate Professor in Statistics (2004-2013), University of Copenhagen.
Visiting Scholar (2010-2011), Biostatistics, University of Berkeley.
2016-2019. Søren Wengel Mogensen. Causality in continuous-time dynamical systems.
2014-2017. Frederik Riis Mikkelsen. Constrained estimation of parameters in nonlinear ODEs.
2013-2016. Adam Lund. Spatio-temporal modeling of neuron fields.
2010-2014. Alexander Sokol. On martingales, causality, identifiability and model selection.
Awarded the Science PhD Award 2014 as one of three outstanding PhD-projects at Science, University of Copenhagen.
2009-2013. Martin Vincent. High dimensional multiclass classification with applications to cancer diagnosis.
2006-2010. Lisbeth Carstensen. Hawkes processes and combinatorial transcriptional regulation.
2015-2018. Editor, Scandinavian Journal of Statistics.
2013- Member of the editorial board of the International Journal of Biostatistics.
2012-2016. Member of the European Regional Committee of the Bernoulli Society (currently Program Coordinator).
2014-2015 Associate Editor, Scandinavian Journal of Statistics.
2008-2010. Chairman of the Danish Society for Theoretical Statistics (DSTS).
Active on the statistics and machine learning Q&A site CrossValidated and on a few other StackExchange sites.