Marius Kloft

Marius Kloft
Rätsch Research Group
Room: Z-677
Tel: +1 646 888 3394
Email: kloft@cbio.mskcc.org
Mailing Address: MSKCC – Computational Biology Center
1275 York Avenue, Box # 357
New York, NY 10065
Express Mail/Street Address: Zuckerman Research Center
417 East 68th Street – 6th Floor (Z-677)
New York, NY 10065

 

Marius Kloft is interested in the foundations of statistical machine learning in general and particularly in multiple kernel learning and learning from non-i.i.d. observations.

Marius earned his M.S. in Mathematics from U Marburg in 2006 with a thesis in algebraic geometry and his Ph.D. in Computer Science from TU Berlin in 2011, where he was advised by  Klaus-Robert Müller. He was co-advised by Gilles Blanchard (U Potsdam), and by Peter L. Bartlett (UC Berkeley) during his one-year research visit at UC Berkeley from Fall 2009 to Fall 2010.  Since Dec 2012 he is a joint postdoctoral fellow at Courant Institute & MSKCC, working with Mehryar Mohri and Gunnar Rätsch, respectively.

Marius had been working on several aspects of multiple kernel learning: (non-sparse) regularization strategies, generalization bounds, unified framework, and novelty detection. He has co-organized workshops on new directions in multiple kernel learning and multi-task learning at NIPS 2010 and 2013, respectively. His dissertation on Lp-norm multiple kernel learning was nominated by TU Berlin for the Doctoral Dissertation Award of the German Chapter of the ACM (GI).

 

Selected Publications:

  1. C. Cortes, M. Kloft, M. Mohri. Learning Kernels Using Local Rademacher Complexity. Advances in Neural Information Processing Systems (NIPS) 26. Spotlight paper. In press. PDF.
  2. X. Lou, M. Kloft, G Rätsch, F. A. Hamprecht. Structured Learning from Cheap Data. In Advanced Structured Prediction. The MIT Press. In press.
  3. C. Widmer, M. Kloft, G. Rätsch. Multi-task Learning for Computational Biology: Overview and Outlook. In B. Schoelkopf, Z. Luo, and V. Vovk, editors, Empirical Inference – Festschrift in Honor of Vladimir N. Vapnik. In press.
  4. C. Widmer, M. Kloft, N. Görnitz, G. Rätsch. Effient Training of Graph-Regularized Multitask SVMs.  Proceedings of the European Conference on Machine Learning and Knowledge Discovery in Databases (ECML PKDD), pages 633-647, 2012.
  5. M. Kloft and P. Laskov. Security Analysis of Online Centroid Anomaly Detection. Journal of Machine Learning Research (JMLR), 13(Dec):3647-3690, 2012. PDF.
  6. M. Kloft and G. Blanchard. On the Convergence Rate of Lp-Norm Multiple Kernel Learning. Journal of Machine Learning Research (JMLR), 13(Aug):2465-2502, 2012. PDF.
  7. M. Kloft, U. Brefeld, S. Sonnenburg, and A. Zien. Lp-Norm Multiple Kernel Learning. Journal of Machine Learning Research (JMLR), 12(Mar):953-997, 2011. PDF.