
Xu Chen
Ph.D. student at UESTC working on machine learning for query optimization and autonomous database agents.
I am a third-year Ph.D. student in the School of Computer Science and Engineering at the University of Electronic Science and Technology of China, advised by Prof. Kai Zheng and Han Su. My research connects machine learning, query optimization, and database agents, with a focus on building autonomous systems that are efficient, practical, and ready for real products.
Research Interests
I study how learning-based methods can improve database optimization and make agentic systems more useful in vertical domains.
- Machine learning for query optimization: combining rule-based optimization with learned components to build efficient, autonomous, and practical database optimizers.
- LLM and database agents: using reinforcement learning and test-time scaling to improve reasoning, tool use, and autonomous database operations.
Publications
Recent Work
Optimizing Block Skipping for High-Dimensional Data with Learned Adaptive Curve.
SIGMOD, 2025.
DACE: A Database-Agnostic Cost Estimator.
ICDE, 2024.
LEON: A New Framework for ML-Aided Query Optimization.
VLDB, 2023.
BASE: Bridging the Gap between Cost and Latency for Query Optimization.
VLDB, 2023.
Earlier Work
Efficient Learning with Pseudo Labels for Query Cost Estimation.
CIKM, 2022.
HeGA: Heterogeneous Graph Aggregation Network for Trajectory Prediction in High-Density Traffic.
CIKM, 2022.
RISE: A Velocity Control Framework with Minimal Impacts based on Reinforcement Learning.
CIKM, 2022.
Astral: An Autoencoder-Based Model for Pedestrian Trajectory Prediction of Variable-Length.
DASFAA (Workshops), 2022.
TSummary: A Traffic Summarization System Using Semantic Words.
DASFAA (Workshops), 2022.
Efficient Join Order Selection Learning with Graph-based Representation.
KDD, 2022.
PATROL: A Velocity Control Framework for Autonomous Vehicle via Spatial-Temporal Reinforcement Learning.
CIKM, 2021.
SCSG Attention: A Self-centered Star Graph with Attention for Pedestrian Trajectory Prediction.
DASFAA, 2021.