First, it knocked down Li Shishi, Ke Jie’s “Alpha Dog”, the top player in the human chess world, and then “zero-based” self-taught, “Alpha” to defeat Alpha Dog 100 to 0. With the birth of "Alfa Dog" and "Alpha Dollar", a round of intensified learning technology has been set up. This direction has become one of the hottest directions in artificial intelligence. Hao Jianye, an associate professor of the software college of Tianjin University, who was selected as the “China Computer Association Youth Talent Supporting Project” in 2017-2019 and an expert of the Tianjin Young Talents Program, has a lot of achievements in the field of deepening reinforcement learning and has successively participated in many related aspects. National and provincial and ministerial-level scientific research projects, and is committed to the implementation of those grand artificial intelligence to the actual daily life.
Depth-enhanced learning is the use of deep learning with the ability to automatically extract features of the network dynamic scene, and then make the best decision through reinforcement learning with decision-making capabilities. Hao Jianye's team improved the performance and efficiency of learning optimal decision-making in a complex environment by researching and designing single-body and multiple-body (depth) reinforcement learning and game algorithms and models. Its research results are mainly used in areas such as automatic negotiation and smart grid.
In the field of smart grids, Hao Jianye team worked with the Imperial College of Technology to design electricity price pricing strategies based on deepening learning to optimize the power market revenue for London's electricity production and consumption data over the years. His research direction was supported by projects supported by the National Natural Science Foundation of China, the Tianjin Natural Science Foundation, the Hong Kong Research Grants Agency, and the Australian Ministry of Education Scholarship. Its research results in the field of e-commerce auto-negotiation: "AbiNes: An Adaptive Bilateral Negotiation Algorithm" won the 2012 International Top Auto-Negotiation Champion; its "Mercury Algorithm" won the 2015 runner-up. At present, the team also cooperates with the NetEase game development department to develop high-efficiency game agents and deepen the user's gaming experience through deep reinforcement learning algorithms for their game products. For example, in the competition game, the artificial intelligence body can show different skill levels according to the level of different game players, and accordingly improve their skills as the level of the user's game increases.
With the advent of the "Internet Plus" era, cyberspace security defenses have become particularly important. The project “Mobile Internet Internet Attack Detection in Big Data Environment” of Hao Jianye, Associate Professor of the School of Software, Tianjin University, proposes a secure game theory model and combines multi-agent reinforcement learning techniques to study and design an optimal defense strategy for dealing with different types of cyber attacks.
A man-in-the-middle attack is a common method of network attack. It means that an attacker establishes a separate connection with the original normal communication by using some technical means. This can be used as a middleman to monitor the entire communication process without the knowledge of the two parties. In the case of stealing sensitive information from the user. Because the existing man-in-the-middle attack defense technology can not completely eliminate the existence of the attack, the previous defense strategy is to prevent the attack by switching ports or encryption. Hao Jianye's team studied the man-in-the-middle attack defense problem from a brand-new perspective: by leveling user information, inducing attackers to attack relatively minor information, by confusing attackers and reducing their probability of acquiring user-sensitive core information. Its related research results "Defense of the Middleman Attack under Repeated Game" and "Optimal Individualized Defense Strategy against the Man-in-the-Middle Attack" were published at the 26th and 31st Artificial Intelligence Conference at the top conference of the artificial intelligence conference.
Aluminum LED Strip,LED Strip,LED Flexible Strip
LED Power Supply Co., Ltd. , http://www.szledstrip.com
没有评论:
发表评论