SPEAKERS
![]() |
Prof. Dr. Naoyuki Kubota Tokyo Metropolitan University Professor in the Department of Mechanical Systems Engineering, the Graduate School of Systems Design. Director of Community-centric Systems Research Center, Tokyo Metropolitan University, Japan. |
Biography
Prof. Dr. Naoyuki Kubota is currently a Professor in the Department of Mechanical Systems Engineering, the Graduate School of Systems Design, and Director of Community-centric Systems Research Center, Tokyo Metropolitan University, Japan. He is the representative director of the Tokyo Biomarker Innovation Research Association, Japan. He received a doctoral degree from Nagoya University, Japan, in 1997. He joined Osaka Institute of Technology and Fukui University, Japan. He was an Associate Professor from 2005 to 2012, and a Professor from 2012 at the Graduate School of Systems Design, Tokyo Metropolitan University, Japan. He was a Visiting Professor at University of Portsmouth, UK and Seoul National University, and others. His current interests are in the fields of topological mapping, coevolutionary computation, spiking neural networks, robot partners, and informationally structured space. He has published more than 500 refereed journal and conference papers in the above research fields. He was an associate editor of the IEEE Transactions on Fuzzy Systems from 1999 to 2010, the IEEE CIS Intelligent Systems Applications Technical Committee, Robotics Task Force Chair from 2007 to 2014, IEEE Systems, Man, and Cybernetics Society, Japan Chapter Chair from 2018 to 2021, IEEE Transactions on Affective Computing Steering Committee Member since 2019, and others. |
ABSTRACT
“Data-X Approach in Machine Learning”
Recently, data-driven approaches have been used widely in big data analysis. For example, a data-driven approach is used directly for end-to-end machine learning without feature space design. The approach is an interpolation method to get the machine understanding within the given data and the resulting model is often considered as a black box. On the other hand, data-informed approach is used for machine learning as explainable artificial intelligence to obtain a white box model. The approach is an extrapolation method to gain the human understanding and general knowledge from the given data. However, if we try to use all the data given, we may only get general findings or conclusions, but if we use a small or partial set of data, we may be able to come up with a different assumption or hypothesis. Therefore, human inspiration to select a small or partial set of data is very important to find new knowledge. Such a human-like method is called a data-inspired approach. In this talk, we discuss the above Data-X approaches used for machine learning in topological intelligence. Topological intelligence is used for inference, learning, search, and prediction based on topological and graphical data extracted from big data or measurement data. First, we discuss the role of topological mapping. Then, we explain different types of topological mapping methods, unsupervised learning methods, and graph-based methods as the methodology of topological intelligence. Furthermore, we show some experimental results of topological intelligence used within the framework of Data-X approaches. Finally, we discuss the future direction of research on the data-inspired approach.