Gunnar Rätsch, Ph.D.

Memorial Sloan-Kettering Cancer Center
415 E 68th street, Room Z690
New York City, NY 10065

Email: ratschg@mskcc.org

Phone: +1 646 888 2802

I earned my PhD at Fraunhofer First and the computer science department of the University of Potsdam. My thesis on Robust Boosting and Convex Optimization can be downloaded here. I was a member of the IDA group under Klaus-Robert Müller. Now I am the group leader of the Machine Learning in Genome Biology group at the Friedrich Miescher Laboratory in Tübingen.
I have worked on Boosting type algorithms and SVMs. But I also find online-learning, optimization theory, new inference principles and new types of machine learning problems very interesting. Furthermore, I am very much interested in applying machine learning real world problems, e.g. from computational biology and chemistry. To find publications by me, checkout Boosting.org and Kernel-Machines.org or my publications page
Recently, I got interested in modern Machine Learning techniques suitable for the analysis of problems arising in Genome Biology, in particular novel methods for ab initio gene finding in nematode genomes and the prediction and validation of transcriptional regulation (e.g. alternative splicing).

Please check our my old home pages at

I take more or less care of the following web sites:

Please also check:

Large Scale Learning with String Kernels:
Boosting:: Before I started my work on computational Biology, I was working on Boosting related learning algorithms. Here, I was particularly interested in the analysis of convergence properties of such algorithms, their relation to large margin learning and to Support Vector Machines.
Efficient Algorithms for Structured Output Learning:
Extracting Discriminative Motifs Based on SVM Detectors:
Domain Adaption and Multi-task Learning:
Spliced Alignments for ESTs, cDNAs, and Next-Generation Sequencing Reads:
Analysis of Whole-Genome Tiling Microarrays:
RNA Secondary Structure Prediction:
Inference of Alternative Splicing:

cBio@MSKCC