Ryosuke Kojima, Ph.D.

Department of Biomedical Data Intelligence,
Graduate School of Medicine,
Kyoto University
Email: kojima.ryosuke.8e[at]kyoto-u.ac.jp

I am studying artificial intelligence (AI) and machine learning technology to apply them to real-world problems. My interests include time-series data processing. There are no established technology to analyze time-series data due to many domain-specific dificulties. I am addressing various real world problems in order to discover and establish an efficient methodology of time series data.


  • Open-sourced robot audition software :HARK
    • used in [t2][c1] for sound source localization and separation
    • I'm a lecturer in the 11th, 12th, and 13th HARK seminars and 4th HARK hackathon
  • A probabilistic programming language:PRISM (github page)
    • used in [t1] for preference learning for and knowledge graph
    • used in [t2] for probability computations on the hierarchical hidden Markov model (HHMM)
    • used in [t3][t4] for the prefix probability computation on the probabilistic context-free grammar (PCFG)
    • I'm a developer (ver. 2.2 or later and T-PRISM(tensorized version of PRISM))
  • A graph neural network platform for life science: kGCN
  • “introduction of machine learning” in KD DHIEP Program (Kyoto University and Deloitte Data-Driven Healthcare Innovation Evangelist promotion Program)


  • 2012: received a B.E. in computer science in 2012 from the Tokyo Institute of technology
  • 2014: received an M.E. in information science and engineering in 2014 from the Tokyo Institute of technology
  • 2017: received a Ph.D.(engineering) in information science and engineering in 2017 from the Tokyo Institute of technology
  • 2017: currently a program-specific associate professor (AMED) at Kyoto University


Conference and workshop

Other conference and workshop


  • Kojima, R. and Sato, T.:ILP 2014 Best Student Paper runner-up
In , , , , () : ed., , .