Sören R. Künzel is currently a fourth-year Ph.D. candidate in the Department of Statistics at UC Berkeley, and he is jointly supervised by Peter Bickel, Jasjeet Sekhon, and Bin Yu. After studying Mathematics and Medizin at the University of Bonn, he went for one year to the Department of Statistics at Yale where he did some research on model selection criterions.
He is very interested in Causal Inference, Machine Learning and Experimental Design, he enjoys solving real world problems, and analyzing asymptotic behavior of statistical estimators. Together with his supervisors and Allen Tang, he has developed an R package to estimate heterogeneous treatment effects (https://github.com/soerenkuenzel/hte), and he is currently developing a new version of random forests which is in particularly well suited for statistical inference.
In addition to this, he is about to release a new algorithm to optimally assign units to different treatment groups using a variation of the Knowledge Gradient criterion applied to Gaussian Process Priors.
When he is neither coding nor reading research papers, then he is probably practicing a new routine on the parallel bars.