Jilin Hu
Professor
School of Data Science and Engineering, East China Normal University
Room 217 at Dili Building, 3663 North Zhongshan Road, Shanghai, China
jlhu [at] dase.ecnu.edu.cn
Google scholar | DBLP
Jilin Hu is a Professor at the School of Data Science and Engineering, East China Normal University. Prior to that, he was an Associate Professor at the Department of Computer Science, Aalborg University. He was a Research Assocaite at Inception Institute of Artificial Intelligence (UAE) under the supervision of Prof. Jianbing Shen. He obtained the Ph.D. degree from Aalborg Unviersity in 2019, under the supervision of Prof. Christian S. Jensen and Prof. Bin Yang. From Oct. 2017 to April. 2018, he visited University of California, Berkeley under the supervision of Prof. Alexandre Bayen.
Research interest: Spatio-Temporal Data Management, Traffic Data Analysis, AI For Science
Never stop looking for highly self-motivated students.
News
| Jan 14, 2026 | One paper “TimeMar: Multi-Scale Autoregressive Modeling for Unconditional Time Series Generation” is accepted by WWW 2026, and congrats to Xiangyu and Qingsong! |
|---|---|
| Jan 14, 2026 | One paper “FSDI: Frequency-Shaped Diffusion For Time-Series Imputation” is accepted by WWW 2026, and congrats to Wangmeng! |
| Dec 17, 2025 | One paper “Dynamic gradient-optimized softmax for efficient transformer acceleration” is accepted by KBS, and congrats to Kai Zhang! |
| Nov 8, 2025 | Our paper “Beyond Dynamic Quantization: An Efficient Static Hierarchical Mix-precision Framework for Near-Lossless LLM Compression” receives EMNLP Best Paper Award (Industry Track)! |
| Nov 8, 2025 | One paper “Spatial-Temporal Feedback Diffusion Guidance for Controlled Traffic Imputation” is accepted as Oral by AAAI 2026, and congrats to Xiaowei Mao and Huihu Ding! |
| Nov 8, 2025 | One paper “SculptDrug: A Spatial Condition-Aware Bayesian Flow Model for Structure-based Drug Design” is accepted as Oral by AAAI 2026, and congrats to Qingsong! |
| Nov 8, 2025 | One paper “DiffMM: Efficient Method for Accurate Noisy and Sparse Trajectory Map Matching via One Step Diffusion” is accepted as Poster by AAAI 2026, and congrats to Chenxu Han! |
| Nov 8, 2025 | One paper “Rethinking Irregular Time Series Forecasting: A Simple yet Effective Baseline” is accepted as Oral by AAAI 2026, and congrats to Xuyuan Liu! |
Selected Publications
(# Corresponding Author)
- AAAI’26SculptDrug: A Spatial Condition-Aware Bayesian Flow Model for Structure-based Drug DesignIn AAAI 2026
- AAAI’26DiffMM: Efficient Method for Accurate Noisy and Sparse Trajectory Map Matching via One Step DiffusionIn AAAI 2026
- AAAI’26Rethinking Irregular Time Series Forecasting: A Simple yet Effective BaselineIn AAAI 2026
- NeuriPS’25DBLoss: Decomposition-based Loss Function for Time Series ForecastingIn NeuriPS 2025
- PVLDB’25TAB: Unified Benchmarking of Time Series Anomaly Detection MethodsIn PVLDB 2025
- KDD’25SSD-TS: Exploring the potential of linear state space models for diffusion models in time series imputationIn KDD 2025
- IJCAI’25ADFormer: Aggregation Differential Transformer for Passenger Demand ForecastingIn IJCAI 2025
- KDD’25DUET: Dual Clustering Enhanced Multivariate Time Series ForecastingIn KDD 2025
- WWW’25Path-LLM: A Multi-Modal Path Representation Learning by Aligning and Fusing with Large Language ModelsIn WWW 2025
- KBSSparseLight: Dynamic gradient-optimized softmax for efficient transformer accelerationKnowledge-Based Systems 2026
- ISsResidual memory inference network for regression tracking with weighted gradient harmonized lossInformation Sciences 2022
- TNNLS