H
Haifeng Zhang
Researcher at Peking University
Publications - 29
Citations - 358
Haifeng Zhang is an academic researcher from Peking University. The author has contributed to research in topics: Reinforcement learning & Computer science. The author has an hindex of 7, co-authored 23 publications receiving 145 citations. Previous affiliations of Haifeng Zhang include University College London.
Papers
More filters
Journal ArticleDOI
Learn to Navigate: Cooperative Path Planning for Unmanned Surface Vehicles Using Deep Reinforcement Learning
TL;DR: This work investigates the application of deep reinforcement learning algorithms for USV and USV formation path planning with specific focus on a reliable obstacle avoidance in constrained maritime environments.
Journal ArticleDOI
Bi-level Actor-Critic for Multi-agent Coordination
TL;DR: This paper proposes a novel bi-level actor-critic learning method that allows agents to have different knowledge base (thus intelligent), while their actions still can be executed simultaneously and distributedly and considers Stackelberg equilibrium as a potentially better convergence point than Nash equilibrium in terms of Pareto superiority.
Proceedings ArticleDOI
Improving Knowledge Tracing via Pre-training Question Embeddings.
TL;DR: It is demonstrated that large gains on KT can be realized by pre-training embeddings for each question on abundant side information, followed by training deep KT models on the obtained embeddeddings.
Proceedings ArticleDOI
User Response Learning for Directly Optimizing Campaign Performance in Display Advertising
TL;DR: This paper reformulates a common logistic regression CTR model by putting it back into its subsequent bidding context: rather than minimizing the prediction error, the model parameters are learned directly by optimizing campaign profit.
Journal ArticleDOI
Offline Pre-trained Multi-agent Decision Transformer
Linghui Meng,Muning Wen,Chenyang Le,Xiyun Li,Dengpeng Xing,Weinan Zhang,Ying Wen,Haifeng Zhang,Jun Wang,Yaodong Yang,Bo Xu +10 more
TL;DR: In this paper , a multi-agent decision transformer (MADT) is proposed to learn generalizable policies that can transfer between different types of agents under different task scenarios, which can improve sample efficiency and generalizability.