Nonlinear System Analysis and its Applications to Machine Learning
My original research field is the ergodic theory of chaos which originates from the foundation of statistical mechanics. How do probabilistic laws appear from deterministic systems? What kind of mathematical framework provide that justification? Driven by such questions, I have studied some fundamental mathematical aspects of nonlinear systems generating chaos. After receiving a doctor's degree, I have taken part in several mathematical studies about Life Science and Information-and-Communication Engineering, making use of my experiences of nonlinear system analysis, and watched various applied mathematical studies. Then I was gradually becoming interested in solving real-world problems by mathematical methods.
The above experiences of nonlinear system analysis are considered to be useful even in the field of Machine Learning. For example, Neural Network, which is a fundamental building block of Deep Learning, is a nonlinear system that has studied in the statistical mechanics for a long time, and Reservoir Computing, which is attracting many researchers in recent years, utilizes various nonlinear systems as reservoirs. Under such backgrounds, I am tackling several fundamental studies and its applications in the field of Machine Learning, making good use of knowledge and my experiences of nonlinear system analysis.
- April 2012: Research Associate at Waseda University
- March 2015: Doctor of Science from Waseda University
- April 2015: Researcher at Hiroshima University
- April 2016: Assistant Professor (Special Appointment) at RcMcD in Hiroshima University
- April 2017: Researcher at The University of Electro-Communications
- April 2019: Specially-appointed Assistant Professor at Osaka University
The four-digit phone numbers are extensions at Osaka University. To call directly from outside Osaka University, dial a phone number marked S after (area code 06) 6850 or a phone number marked T after (area code 06) 6879. For a phone number marked S (extension), however, first dial the main phone number (06) 6879-5111.
The e-mail addresses are given without the suffix "osaka-u.ac.jp"; add "osaka-u.ac.jp" to each e-mail address.