List of faculty
members

FUJISAWA, Katsuki( FUJISAWA, Katsuki ) / Professor

Faculty , Department

Institute of Mathematics for Industry : Intelligent Societal Implementation of Mathematical Computation ,Laboratory of Mathematical Design for Advanced Cryptography (concurrent) ,Industrial and Mathematical Statistics (concurrent)

Keywords

mathematical optimization, graph analysis, high-performance computing, machine learning, deep learning

Large-scale graph analysis and data processing are attracting attention as new supercomputer applications. Possible applications of graph analysis include evacuation planning for large-scale disasters, effective use of large-scale data such as social networks for social and public policy and corporate management, etc. However, the scale of these applications is extremely large in terms of computation, data volume, and power consumption, making them difficult to process using conventional methods. Therefore, we are researching to realize ultra-large-scale graph processing using techniques in the field of high-performance computing. In 2015 (ISC15, June 2015), we achieved the world's highest result in Graph500, which measures big data processing on supercomputers by solving large-scale graphs using our originally developed software. The research team has been working on a domestic supercomputer. The research team used domestic supercomputers (supercomputers Fugaku and K computer) to achieve excellent results, including the world's first place on supercomputers Fugaku and K computer (8th, 10th to 18th, and 20th to 23rd Graph500 benchmarks).
The semi-positive definite programming problem (SDP), for example, has a wide range of applications in combinatorial optimization, systems and control, data science, financial engineering, and quantum chemistry and is currently one of the most popular optimization problems in the field of optimization research. Furthermore, with large-scale parallel computing on supercomputers, we have succeeded in solving the world's largest SDP in a very fast time. The department is now actively promoting research to solve complex and unsolved problems in the real-world using mathematics and computers in collaboration with other research institutes and companies. For these accomplishments, we received the Prizes for Science and Technology (Research Category) of the Commendation for Science and Technology by the Minister of Education, Culture, Sports, Science and Technology in 2017, among others.
In recent years, various efforts to realize a secure, safe, and convenient so-called "super-smart society" (Society 5.0, etc.) by combining and integrating the latest technologies have been promoted around the world. Recent improvements in ICT have made it possible to realize a cyber-physical system (CPS) as a business model by modeling phenomena occurring in the real world in advance on a computer and then simulating and optimizing them in response to environmental changes. Currently, in collaboration with many private companies, we are developing a CPS mobility optimization engine (CPS-MOE) that performs optimization, machine learning, deep learning, and simulation in cyberspace using a large amount of sensor data (movement of people and goods) and open data (movement history of Wi-Fi, etc.) for the CPS. To realize the CPS-MOE, we propose and develop new mathematical and information technologies to represent, predict, optimize, and control mobility, especially in the following three areas. We are promoting the proposal and development of new mathematical and information technologies to express, predict, optimize, and control the following three mobilities in particular
1: Mobility of information (human interest and intention): user clustering using web access movement data and user's potential interest
2: Mobility of people and things: location detection and tracking (deep learning), congestion detection and flow optimization and visualization
3: Mobility for transportation (optimal driving): route optimization, delivery optimization, MaaS (bike sharing, etc.)