Hi! I’m Zuoyu Yan (严祚宇), a final-year Ph.D. student at Peking University, advised by Prof. Zhi Tang and Prof. Liangcai Gao. Before starting Ph.D., I finished my undergraduate study at the Yuanpei College of Peking University, an exemplary high-quality and pioneering undergraduate college, majoring in Data Science.

During my Ph.D., I have also been fortunate to spend time collaborating with Prof. Chao Chen, Prof. Yusu Wang, Prof. Tengfei Ma, and Prof. Muhan Zhang.

Research Interests: My research interest lies in the intersection of machine learning and topological data analysis. In particular, I am interested in:

1. Enhancing modern machine learning frameworks, e.g., graph neural networks, with topology/geometry.
2. Applying these models in real-world scenarios, e.g., biomedicine, and recommendation systems.

This is my CV. If you are interested in my research, don’t hesitate to get in touch with me through yanzuoyu3@pku.edu.cn.

📝 Selected Publications

• Cycle Invariant Positional Encoding for Graph Representation Learning [Paper][Code]

  Zuoyu Yan, Tengfei Ma, Liangcai Gao, Zhi Tang, Chao Chen, Yusu Wang

  The Second Learning on Graphs Conference (LoG), 2023 (Oral)

• Efficiently Counting Substructures by Subgraph GNNs without Running GNN on Subgraphs [Paper] [Code]

  Zuoyu Yan, Junru Zhou, Liangcai Gao, Zhi Tang, Muhan Zhang

  aiXiv Preprint.

• Neural Approximation of Graph Topological Features [Paper] [Code]

  Zuoyu Yan, Tengfei Ma, Liangcai Gao, Zhi Tang, Yusu Wang, Chao Chen

  Thirty-sixth Conference on Neural Information Processing Systems (NeurIPS), 2022 (Spotlight)

• Cycle Representation Learning for Inductive Relation Prediction [Paper] [Code]

  Zuoyu Yan, Tengfei Ma, Liangcai Gao, Zhi Tang, Chao Chen

  International Conference on Machine Learning (ICML), 2022

• Link Prediction with Persistent Homology: An Interactive View [Paper] [Code]

  Zuoyu Yan, Tengfei Ma, Liangcai Gao, Zhi Tang, Chao Chen

  International Conference on Machine Learning (ICML), 2021

• Handwritten Mathematical Expression Recognition with Bidirectionally Trained Transformer [Paper] [Code]

  Wenqi Zhao, Liangcai Gao, Zuoyu Yan, Shuai Peng, Lin Du, Ziyin Zhang

  International Conference on Document Analysis and Recognition (ICDAR), 2021 (Best Poster Award, 2 out of 340 submissions)

🏫 Professional Services

Conference Reviewer for:

• Conference on Neural Information Processing Systems (NeurIPS)
• International Conference on Machine Learning (ICML)
• International Conference on Learning Representations (ICLR)
• Learning on Graphs Conference (LoG)