example cover photo

Hi ! We are a team of
Deep Reinforcement Learning .
Welcome to join our team !

About Our Group

Following the stunning success of AlphaGo, Deep Reinforcement Learning (DRL) combining deep learning and conventional reinforcement learning has emerged as one of the most competitive approaches for learning in sequential decision making problems. It links with the advanced machine learning theory, optimization theory and statistical theory, which sets a complex task for the students who are interested in this field.

Driven/inspired by many talented researchers, our 1029DRL Group (led by [礼欣老师] [A/Prof. Xin Li] keeps investigating the emerging DRL algorithms published in the top conferences, aiming to develop our own to contribute to the RL community. We devote to developing a more explainable/sample-efficient/training-efficient/robust (D)RL algorithm with applications to single/multi-player games, robotics and healthcare.

We are currently recruiting Ph.D candidates, master students, senior undergradate students to join our group to work with us and share the thoughts.

Bulletin
  • 05/2024: Congratulations to Ruixiang Sun and Hongyu Zang [Learning Latent Dynamic Robust Representations for World Models] got accepted for ICML 2024.
  • 12/2023: Congratulations to Min Wang [MetaCARD: Meta-Reinforcement Learning with Task Uncertainty Feedback via Decoupled Context-Aware Reward and Dynamics Components] got accepted for AAAI 2024.
  • 12/2023: Congratulations to Yujie Fang [Improving GNN Calibration with Discriminative Ability: Insights and Strategies] got accepted for AAAI 2024.
  • 09/2023: Congratulations to Hongyu Zang [Understanding and Addressing the Pitfalls of Bisimulation-based Representations in Offline Reinforcement Learning] got accepted for NeurIPS 2023.
  • 06/2023: Congratulations to Peiyao Zhao [Coarse-to-Fine Contrastive Learning on Graphs] got accepted for TNNLS 2023.
  • 06/2023: Congratulations to Haojie Lei [Differentiable Logic Policy for Interpretable Deep Reinforcement Learning: A Study from an Optimization Perspective] got accepted for TPAMI 2023.
  • 10/2022: Our collaborative paper [Agent-Controller Representations: Principled Offline RL with Rich Exogenous Information] got accepted for ICML 2023.
  • 01/2023: Congratulations to Hongyu Zang. Our paper [Behavior Prior Representation learning for Offline Reinforcement Learning] got accepted for ICLR 2023.
  • 01/2023: Congratulations to Hongyu Zang. Collaborating with Mila and MSR, our paper [Representation Learning in Deep RL via Discrete Information Bottleneck] got accepted for AISTATS 2023.
  • 12/2022: Congratulations to Qianyu Chen and Fuhao Yang, two papers [Context-aware Safe Medication Recommendations with Molecular Graph and DDI Graph Embedding] and [WaveForM: Graph Enhanced Wavelet Learning for Long Sequence Forecasting of Multivariate Time Series] got accepted for AAAI 2023 oral.
  • 11/2022: Congratulations to Hongyu Zang. A collaborative paper [Discrete Compositional Representations as an Abstraction for Goal Conditioned Reinforcement Learning] got accepted for NeurIPS 2022.
  • 12/2021: Congratulations to Hongyu Zang. Our paper[SimSR: Simple Distance-based State Representation for Deep Reinforcement Learning] got accepted for AAAI 2022 oral.
  • 09/2021: Congratulations to Zhang Li. Our paper[Off-Policy Differentiable Logic Reinforcement Learning] got accepted for ECML-PKDD 2021.
  • 05/2020: Another work related to POMDP tasks [On Improving the Learning of Long-Term historical Information for Tasks with Partial Observability] is published.
  • 04/2020: Collaborating with Ubisoft, we organized some competitions to test our RL agent in Rabbids: Journey To The West, and our RL agent won all the games.
  • 02/2020: Zhang Li gives the oral presentation of our UVIN paper in [a recorded video] at AAAI2020 due to the COVID-19 pandemic.
  • 12/2019: The students won the best technical award in the Data Hackathon competition hosted by Ubisoft.
  • 12/2019: Congratulations to Zhang Li. Our [UVIN paper] got accepted for AAAI 2020 oral presenations.
  • 04/2017: The group submit our ADRQN paper named [On Improving Deep Reinforcement Learning for POMDPs] to Arxiv (cited over 104 times).