Biography

I am a CS Ph.D. student at the University of Michigan, Ann Arbor, advised by Prof. Lin Ma. My research lies at the intersection of machine learning and data systems. I currently work on learned query optimization across heterogeneous database systems, with an emphasis on building methods that are accurate, efficient, and reusable across different database engines. I also study automated optimization for lakehouse systems, including the joint selection and configuration of execution engines and table formats. More broadly, I am interested in intelligent and portable systems for data management.

I finished my Bachelor degree in Computer Science and Technology at Peking University, where I worked on approximate algorithms in data streams and computer network systems with Prof. Tong Yang. I also finished my Bachelor degree in Economics at Peking University.

Publications

  • M4: A Framework for Per-Flow Quantile Estimation
  • Siyuan Dong, Zhuochen Fan, Tianyu Bai, Tong Yang, Hanyu Xue, Peiqing Chen and Yuhan Wu.
    2024 IEEE 40th International Conference on Data Engineering (ICDE '24).
  • Kvsagg: Secure aggregation of distributed key-value sets
  • Yuhan Wu, Siyuan Dong (Co-first) , Yi Zhou, Yikai Zhao, Fangcheng Fu, Tong Yang, Chaoyue Niu, Fan Wu, and Bin Cui
    2023 IEEE 39th International Conference on Data Engineering (ICDE '23).
  • MicroscopeSketch: Accurate Sliding Estimation Using Adaptive Zooming
  • Yuhan Wu, Shiqi Jiang, Siyuan Dong (Co-first), Zheng Zhong, Jiale Chen, Yutong Hu, Tong Yang, Steve Uhlig, and Bin Cui.
    2023 ACM 29th Conference on Knowledge Discovery and Data Mining (SIGKDD '23).
  • Unbiased Real-time Traffic Sketching
  • Yuhan Wu, Shiqi Jiang, Yifei Xu, Siyuan Dong, Kaicheng Yang, Peiqing Chen, and Tong Yang.
    2023 IEEE Transactions on Network Science and Engineering.
  • HoppingSketch: More Accurate Temporal Membership Query and Frequency Query
  • Zhuochen Fan, Yubo Zhang, Siyuan Dong, Yi Zhou, Fangyi Liu, Tong Yang, Steve Uhlig, and Bin Cui.
    2022 IEEE Transactions on Knowledge and Data Engineering.