Thomas Bühler

Researcher/Ph.D. Student,
Faculty of Mathematics and Computer Science,
Saarland University


Address:
Building E 1 1, Room 227
Universität des Saarlandes
PO Box 15 11 50
D - 66041 Saarbrücken
Germany

phone: +49 - (0)681 - 302 57332
email: tb (at) cs (dot) uni-saarland (dot) de

About me

Since April 2009 I am a researcher and Ph.D. student in the Machine Learning Group at Saarland University, under supervision of Prof. Matthias Hein. I obtained B.Sc. and M.Sc. degrees in Computer Science from Saarland University in 2007 and 2009, respectively.


Research

My general research interest lies in graph-based methods in machine learning in particular unsupervised learning, as well as applications in image processing and computer vision.

Currently my main focus is on spectral properties of the graph p-Laplacian and its applicability for problems in machine learning. The graph p-Laplacian is a nonlinear generalization of the well-known graph Laplacian which has been successfully applied e.g. in clustering and semi-supervised learning. In our recent ICML paper, we showed that spectral clustering based on the graph p-Laplacian, though computationally more expensive, generally has a superior performance compared to standard spectral clustering (read more...).


Publications

  • S. Rangapuram, T. Bühler and M. Hein
    Towards Realistic Team Formation in Social Networks based on Densest Subgraphs
    accepted at WWW 2013

  • T. Bühler, S. Rangapuram, S. Setzer and M. Hein
    Constrained fractional set programs and their application in local clustering and community detection
    In Proc. 30th International Conference on Machine Learning (ICML 2013), JMLR W&CP 28 (1): 624-632, 2013
    PDF (Supplementary material: PDF),

  • M. Hein and T. Bühler
    An inverse power method for nonlinear eigenproblems with applications in 1-spectral clustering and sparse PCA
    In Advances in Neural Information Processing Systems 23 (NIPS 2010), 847-855, 2010.
    PDF (Supplementary material: PDF). Code.

  • T. Bühler and M. Hein
    Spectral Clustering based on the graph p-Laplacian
    In Proc. 26th International Conference on Machine Learning (ICML 2009), 81-88, Omnipress, 2009.
    PDF (Supplementary material: PDF - Errata of Supp. Mat.: PDF). Code.

  • N. Slesareva, T. Bühler, K. Hagenburg, J. Weickert, A. Bruhn, Z. Karni and H.-P. Seidel
    Robust Variational Reconstruction from Multiple Views
    In Proc. 15th Scandinavian Conference on Image Analysis (SCIA 2007), 173-182, Springer, 2007.
    PDF