About

me

I am Professor of Computational Statistics at the department of statistics at LMU Munich. I like to work in data science, machine learning and everything computational in statistics. nd my offical contact details her

Contact

Institut für Statistik 
Ludwig-Maximilians-Universität München
Ludwigstraße 33
D-80539 München
Room 341, 3rd floor
Phone: +49 89 2180 3165

eMail: bernd.bischl@stat.uni-muenchen.de or bernd_bischl@gmx.net

No specific hours. Either visit me directly or send an email.

News

  • June 2015: Was appointed Professor of “Computational Statistics” at LMU.
  • WS 2014/15: Back in Dortmund again. Substitute professorship “Mathematical Statistics and Applications in Industry” at this department.
  • SS 2014: Substitute professorship “Computational Statistics” at the LMU in Munich.
  • Dec 2013: Was granted best dissertation Award for 2013 by my department.
  • Aug. 2013: Defended my PHD thesis.
  • May 2013: The two student teams Klaus Friedrichs and myself supervised in our case studies course this semester just made first AND second place at the data mining cup competition.

Research interest

  • Machine Learning
  • Support Vector Machines and Kernel Methods
  • Variable Selection
  • Model-based Optimization, Kriging and Expected Improvement
  • Algorithm Selection
  • Algorithm Configuration / Tuning
  • R and Statistical Software Development

You can find me on

Software

I am the author or a major contributor to quite a large number of R packages. Nearly all of them are listed and hosted under either my personal github page. They mainly deal with machine learning, optimization and parallelization.

Major projects:

  • mlr - A large ML toolbox for R.
  • mlrMBO - Model-Based Optimization toolbox.
  • OpenML R Package - OpenML connector in R.
  • BatchJobs - Run R jobs on HPC clusters.
  • BatchExperiments - Structure your experiments nicely, then run them with BatchJobs.
  • parallelMap - The names says it. Works also with BatchJobs.
  • aslib - Algorithm selection repo.
  • aslib-r - aslib connector for R.

Awards

  • Best dissertation award 2013 from Department of Statistics, TU Dortmund.
  • I have regularly taught and lead a student team for the German Data Mining Cup competition. We scored 1st place in 2013 and 2011, and 2nd place in 2010 and 2013.
  • Julia Schiffner, Claus Weihs and I won the best paper “Method Award” of the GfKl 2010 conference for our contribution “Bias-Variance Analysis of Local Classification Methods”.
  • Hong Qinyuan won the Industrial Challenge of the GECCO 2011 conference for his “Dynamic Linear Trading System”. He submitted his solution as a student under my supervision.
  • Best intermediate diploma award in computer science from University of Hamburg.

Membership