Koji Tsuda / Professor / Division of Biosciences
Department of Computational Biology and Medical Sciences / / Bioinformatics

Career Summary
1994: Graduated from the Faculty of Engineering, Kyoto University
1998: Received Ph.D. in Engineering from Kyoto University
1998: Researcher, Electrotechnical Laboratory, Tsukuba
2000: Visiting Researcher, GMD FIRST, Germany
2001: Researcher, AIST Computational Biology Research Center
2003-2004, 2006-2008: Senior Researcher, Max Planck Institute for Biological Cybernetics, Germany
2009: Senior Researcher, AIST Computational Biology Research Center
2014: Professor, University of Tokyo
Educational Activities
Undergraduate: Biological Data Mining
Graduate school: Advanced Biological Data Mining
Research Activities
In the beginning of my career, I studied machine learning theory and proposed several methods including the kernel subspace method, the TOP kernel and marginalized kernels [1][2]. I then proposed matrix exponentiated gradient update for online learning [3] and conducted research on structured data such as graphs and trees [4]. Recently, I have developed methods for detecting and testing combinatorial causes from large scale data [5].

1) K. Tsuda et al., Marginalized kernels for biological sequences, Bioinformatics, 18(Suppl. 1):S268–S275, 2002.
2) K. Tsuda et al., A new discriminative kernel from probabilistic models. Neural Computation, 14(10):2397–2414, 2002.
3) K. Tsuda et al., Matrix exponentiated gradient updates for online learning and Bregman projection, Journal of Machine Learning Research, 6:995–1018, 2005.
4) H. Saigo et al., gBoost: A mathematical programming approach to graph classification and regression, Machine Learning, 75:69-89, 2009.
5) A. Terada et al., Statistical significance of combinatorial regulations, PNAS, 110(32):12996-13001, 2013.

Other Activities
Institute of Electronics, Information and Communication Engineers of Japan (IEICE)
Information Processing Society of Japan (IPSJ)
Japanese Society for Bioinformatics (JSBI)
Future Plan
In future, I would like to develop novel data analysis algorithms that eventually lead to great scientific discoveries.
Messages to Students
In many fields of science, computational algorithms and machine learning methods are indispensable. I hope many of you join us because there are lots of opportunities for young people.