In September 2003, Feng Yunhe was admitted to the Command Automation Engineering Department of the National University of Defense Technology. From 20 1 1 to 20 13, he went to the Statistics Department of Harvard University and the High Performance Computing Laboratory of the University of Iowa for further study and served as an assistant researcher.
On June 20 14, he received his Ph.D. from National University of Defense Technology, PLA, and then stayed on as a teacher. He has been a lecturer, deputy director of the teaching and research section, associate professor, deputy director of a provincial and ministerial laboratory, and doctoral supervisor. In July 2023, 1 died in Beijing at the age of 38.
As the core backbone, he has participated in many major national defense scientific research projects and plans, and is one of the main planners and practitioners in the field of national defense intelligence. He actively contacted well-known local experts and scientific research teams, gathered various forces to serve the national defense cause, made outstanding achievements in the fields of national defense intelligent games, and played an important role in the application of artificial intelligence technology in the national defense field.
achievements in scientific research
Feng Yunhe is mainly engaged in reinforcement learning, intelligent game, intelligent planning, chess deduction technology and Bayesian theory, and has made innovative breakthroughs in command and control and intelligent decision-making.
Aiming at the problem of multi-agent intelligent decision-making in complex confrontation environment, Yun Feng-He proposed to construct a "super brain"-"battle skeleton" to assist decision-making in the environment of war chess deduction.
The research and development of "Warhead" is based on the idea of intelligent system engineering, which integrates a series of methods such as knowledge reasoning, supervised learning, semi-supervised learning, integrated learning and reinforcement learning to build an intelligent decision-making model, and makes full use of high-quality data generated by everyone's confrontation and big data generated by machines to train agents.
It realizes the close cooperation of all elements, efficient information processing and accurate and rapid decision-making, and solves the problems of traditional task planning experts, such as strong dependence on experience, weak ability to deal with incomplete information and difficult dynamic adjustment. By July 2023, Feng Yunhe had published more than 60 papers and 4 books, which were cited 1000 times.