cv
Contact Information
| Name | Hanyan Yin |
| Professional Title | |
| yinhanyan@ruc.edu.cn | |
| Website | https://yinhanyan.github.io/ |
Professional Summary
Experience
-
2026 - Present -
2022 - 2023 R&D Intern
Momenta
Participated in the development of 3D map tile graphics compiler
- 3D map tile graphics compiler
Education
-
2025 - Present Beijing, China
-
2023 - 2025 Beijing, China
Master
Gaoling School of Artificial Intelligence, Renmin University of China
Artificial Intelligence
-
2019 - 2023 Beijing, China
B.E.
School of Computer Science, Beijing University of Posts and Telecommunications
Computer Science and Technology
Awards
-
2026 -
2024 First-Class Scholarship for Excellence in Graduate Studies
Renmin University of China
-
2024 -
2023 Outstanding Undergraduate Student of Beijing
Beijing Municiplal Education Commission
-
2023 Outstanding Graduate of Beijing University of Posts and Telecommunications
Beijing University of Posts and Telecommunications
-
2021 CCF-CSP (China Computer Federation Certified Software Professional)
China Computer Federation
Achieved a score of 370, ranked in the top 1.51%.
-
2020 Gold Medal, 14th BUPT Programming Contest (CodeMao Cup)
Beijing University of Posts and Telecommunications
Publications
-
2026 AeroSketch: Near-Optimal Time Matrix Sketch Framework for Persistent, Sliding Window, and Distributed Streams
Proceedings of the ACM on Management of Data (SIGMOD 2026)
The paper introduces AeroSketch, a matrix sketching framework for persistent, sliding window, and distributed streams that achieves optimal communication and space costs with near-optimal update time complexity.
-
2026 Revisiting Matrix Sketching in Linear Bandits: Achieving Sublinear Regret via Dyadic Block Sketching
The Fourteenth International Conference on Learning Representations (ICLR 2026)
The paper proposes Dyadic Block Sketching, a multi-scale matrix sketching approach for linear bandits that achieves sublinear regret without requiring prior knowledge of the streaming matrix properties.
-
2025 Predictive Configuration on DHCP in WLANs
IEEE/ACM Transactions on Networking
The paper proposes PredHCP, a predictive configuration framework that uses deep learning models to improve DHCP lease time and IP pool configuration in WLANs.
-
2024 Optimal Matrix Sketching over Sliding Windows
VLDB 2024 Best Research Paper Nomination
The paper introduces the DS-FD algorithm, which achieves the optimal $O(d/\epsilon)$ space bound for matrix sketching over row-normalized, sequence-based sliding windows, and establishes a space lower bound for time-based and unnormalized sliding windows.