Thomas Gärtner

University of Bonn and Fraunhofer IAIS
Computational Aspects of Mining and Learning
Tel.: +49 2241 14 3513    E-Mail: thomas.gaertner iai.uni-bonn.de
Schloss Birlinghoven
53757 Sankt Augustin
I am leading the research group on computational aspects of mining and learning, jointly carried by the University of Bonn and Fraunhofer IAIS. Currently my main research interest is in constructive machine learning and efficient well-behaved pattern mining through sampling.

I am looking for students that are not afraid of mathematics/theoretical computer science and want to pursue state-of-the-art research in machine learning. Most important is your drive to discover the unknown.

We are organising the 1st international workshop on Constructive Machine Learning @NIPS and look forward to your submissions!

"Our critics are our friends, because they show us our faults." - Benjamin Franklin
"Even if you're on the right track, you'll get run over if you just sit there" - Will Rogers


Overview | Publications | Activities ]


Some selected publications are listed below. I am trying to keep the complete list up to date, though you should also look [@DBLP], [@springer], [@scholar], or [@acm]. Now there is also a list [@microsoft academic search] as well as [@scholar/citations].

Overview | Publications | Activities ]



  • Action Editor since 2009 and Editorial Board Member since 2006 of the Machine Learning Journal ( MLJ )
  • Guest Editor Machine Learning special issue on Mining and Learning with Graphs
  • Tutorials, Colloquia, Invited Talks, ...
    • Learning with Structured Data keynote at the area meeting of the B-IT Research School Area 7 (2011)
    • Machine Learning invited talk at LEARNTEC-2011
    • Kernel Methods for Structured Inputs and Outputs invited talk at the Eighth Workshop on Mining and Learning with GraphsMLG-2010 )
    • Algorithms for Predicting Structured Data tutorial with Shankar Vembu at the European Conference on Machine Learning ( ECML-2010, more details )
    • Computational Aspects of Learning with Graphs lecture at the 2nd CompView Fall School (2009), Japan
    • Machine Learning tutorial at the 2009 workshop of the DFG program Visual Analytics
    • Machine Learning tutorial at the 2008 workshop of the DFG program Organic Computing
    • Kernel Methods in Machine Learning tutorial at the Estonian Summer School in Computer and Systems Science (2006)
    • Kernels for Structured Data at the workshop Learning Structured Information in Natural Language Applications (held with EACL'06)
    • Mining Graph Data with Kernel Methods at the 4th Workshop on Multi-Relational Data Mining (held with KDD'05)
    • Dagstuhl seminar: Probabilistic, Logical and Relational Learning
    • ICML Tutorial: Kernels for Structured Data The application of kernel methods to data that has no natural representation in a Euclidean space, requires the definition of positive definite functions on such data. This tutorial describes the state of the art of a field emerging as the intersection between kernel methods and structured data. The primary goal of this tutorial is to give the audience a coherent overview of the different approaches of applying kernel methods to structured data. It emphasises the conceptual similarities and differences rather than the technical details of the different approaches (more details).
    • An Introduction to Kernel Methods for Mining Graph Data at the research colloquium, TU Berlin.
    • An Introduction to Kernel Methods for Mining Graph Data at the First International Workshop on Mining Graphs, Trees and Sequences (held with ECML/PKDD'03).
    • Bestimmung der Skelettstruktur von Diterpenen durch NMR-Analyse at the biomathematical colloquium, Bonn University. [abstract]
    • Kernel-Based Learning from Structured Data at the 2002 Summer School on Relational Data Mining in Helsinki. [handouts]
  • Programm Committee Membership
  • Research Proposal Evaluation
    • The Israel Science Foundation ( ISF ),
    • Research Foundation Flanders ( FWO )
  • Visits (of varying lengths --- between a few hours and a few months)
    • Xerox Research Centre Europe (2010)
    • IBM TJ Watson Research Center (2010)
    • Laboratory of Computer Sciences, Paris 6 (2010)
    • INSA Rouen (2008)
    • University of Bristol (2008)
    • University of Bristol (11/2007)
    • HIIT Helsinki (10/2007)
    • MPI Tübingen (11/2006)
    • University of Bristol (06/2006)
    • University of Wisconsin (08/2005)
    • National ICT Australia (03-05/2005)
    • TU Berlin (02/2004)
    • University of Freiburg (02/2003)
    • University of Leuven (04/2003)


Overview | Publications | Activities ]

Back to people list