A5101661885- Chinese history and philosophy
- Reinforcement Learning in Robotics
- Robot Manipulation and Learning
- Innovative Teaching and Learning Methods
- Educational Reforms and Innovations
- Impact of Technology on Adolescents
- Social Robot Interaction and HRI
- Regional Economic and Spatial Analysis
- Higher Education and Teaching Methods
- Online and Blended Learning
- Neural Networks and Applications
- Child Development and Digital Technology
- Education and Work Dynamics
- Gender and Technology in Education
- Vietnamese History and Culture Studies
- Online Learning and Analytics
- Advanced Technologies in Various Fields
- Regional Development and Environment
- Speech and dialogue systems
- Innovative Educational Techniques
- Adversarial Robustness in Machine Learning
- Multimodal Machine Learning Applications
- Autonomous Vehicle Technology and Safety
- Japanese History and Culture
- Advanced Memory and Neural Computing
South China Normal University
2023-2024
Hebei Academy of Sciences
2007-2012
Academy of Social Sciences
2009-2012
Czech Academy of Sciences, Institute of History
2009
Trust region methods rigorously enabled reinforcement learning (RL) agents to learn monotonically improving policies, leading superior performance on a variety of tasks. Unfortunately, when it comes multi-agent (MARL), the property monotonic improvement may not simply apply; this is because agents, even in cooperative games, could have conflicting directions policy updates. As result, achieving guaranteed joint where each agent acts individually remains an open challenge. In paper, we extend...
Multi-task learning (MTL) is widely utilized across a variety of real-world applications, including recommendation systems. For instance, in the field e-commerce, MTL commonly employed to simultaneously model click, conversion, and user dwelling time. Among various models, Multi-gate Mixture-of-Experts (MMoE) has gained significant popularity. However, MMoE suffers from polarization issue during training, where weights certain experts tend converge towards 0. To address this issue, we...
Developing reinforcement learning algorithms that satisfy safety constraints is becoming increasingly important in real-world applications. In multi-agent (MARL) settings, policy optimisation with awareness particularly challenging because each individual agent has to not only meet its own constraints, but also consider those of others so their joint behaviour can be guaranteed safe. Despite importance, the problem safe been rigorously studied; very few solutions have proposed, nor a...
The socially-aware navigation system has evolved to adeptly avoid various obstacles while performing multiple tasks, such as point-to-point navigation, human-following, and -guiding. However, a prominent gap persists: in Human-Robot Interaction (HRI), the procedure of communicating commands robots demands intricate mathematical formulations. Furthermore, transition between tasks does not quite possess intuitive control user-centric interactivity that one would desire. In this work, we...
The application of artificial intelligence has benefited vocational English writing courses but still lacks the capacity to dispose contextual challenges. This study compared effectiveness two teaching methods in a course. A method that combines intelligent evaluation and peer review was applied for experimental group, while control group adopted conventional discussion. In this study, quasi-experimental research used. Scores students’ essays as well quality content were examined. results...