Takuo Matsubara
Takuo’s research interests are in Bayesian statistics, statistical machine learning, probabilistic numerics, reproducing Kernel Hilbert space. Their project covers Bayesian quadrature of neural networks.
Project title
'Bayesian quadrature of neural networks'
Research interests
- Bayesian statistics
- statistical machine learning
- probabilistic numerics
- reproducing Kernel Hilbert space
Education
- PhD in School of Mathematics, Statistics and Physics, Newcastle University
- Doctoral Student, The Alan Turing Institute
- MEng in Department of Electrical Engineering and Bioscience, Waseda University
- BEng in Department of Electrical Engineering and Bioscience, Waseda University
Workshop presentations
- Sonoda S., Ishikawa I., Ikeda M., Hagihara K., Sawano Y., Matsubara T., Murata N. “An explicit expression for the global minimizer network“ Poster Presentation at Deep Learning Theory Workshop of The 35th International Conference on Machine Learning, July 2018, Stockholm, Sweden
- Matsubara T., Sonoda S., Murata N. “New Learning Algorithm of Neural Network Using Integral Representation and Kernel Herding” Oral Presentation at The 28th Infomation-Based Induction Sciences and Machine Learning, Mar. 2017, Tokyo, Japan
- Shimada T., Matsubara T., Sonoda S., Murata N., Ortal P., Kato S. “CNN Brake Scene Recognition with LiDAR” Poster Presentation at The 18th Infomation-Based Induction Sciences Workshop, Nov. 2015, Ibaraki, Japan
Supervisors
Contact
Email: tmatsubara@turing.ac.uk