I am a Ph.D. candidate in the Department of Computer Science and Engineering at Shanghai Jiao Tong University (SJTU, 上海交通大学), supervised by Prof. Junchi Yan (严骏驰). I received my bachelor degree from SJTU in 2021, majoring in Computer Science.
My research focuses on deep learning for temporal data modeling, partically time series and event sequences. I have published five independent first-authored paper at top-tier conferences (ICLR-23&26, ICML-24, IJCAI-21&22), with over 2,100 citations. Notably, my ICLR-23 paper, Crossformer, has been cited by AI pioneers including Prof. Yoshua Bengio (NeurIPS-23) and Prof. Jürgen Schmidhuber (ICLR-25).
I serve as a reviewer for leading conferences (NeurIPS 2023-2025, ICML 2023-2026, ICLR 2024-2026) and journals (TPAMI, JMLR, TMLR, TKDE). I was recognized as a Top Reviewer for NeurIPS-24&25
Update: I expect to graduate in June 2027 and am currently seeking Spring/Summer 2026 internship opportunities in LLMs, quantitative trading, and time-series modeling. Feel free to reach out at zhangyunhao@sjtu.edu.cn.
🔥 News
- 2026.01: 🎉🎉 One paper on loss functions for diverse time series forecasting was accepted by ICLR-2026!
📝 Publications
Time Series Modeling
MMPD: Diverse Time Series Forecasting via Multi-Mode Patch Diffusion Loss
Yunhao Zhang, Wenyao Hu, Jiale Zheng, Lujia Pan, Junchi Yan
- A loss for patch-based time series forecasting backbones to model complex future distributions, enabling them to generate multiple diverse predictions with corresponding probabilities.
|
UP2ME: Univariate Pre-training to Multivariate Fine-tuning as
a General-purpose Framework for Multivariate Time Series Analysis
Yunhao Zhang, Minghao Liu, Shengyang Zhou, Junchi Yan
- A general-purpose framework for multivariate time series analysis: univariate pre-training followed by multivariate fine-tuning.
Crossformer: Transformer utilizing cross-dimension dependency for multivariate time series forecasting
Yunhao Zhang, Junchi Yan
-
A Transformer that explicitly utilizes cross-dimension(cross-channel) dependency for multivariate time series forecasting.
-
Event Sequence Modeling
Learning Mixture of Neural Temporal Point Processes
for Multi-dimensional Event Sequence Clustering
Yunhao Zhang, Junchi Yan, Xiaolu Zhang, Jun Zhou, Xiaokang Yang
-
A general framework that mixs multiple Neural Temporal Point Processes (NTTPs) for event sequence clustering.
-
Neural Relation Inference for Multi-dimensional Temporal Point Processes via Message Passing Graph
Yunhao Zhang, Junchi Yan
-
A neural relation inference model for multi-dimensional event sequences that discovers relations among different type of events.
-
Others
-
LinSATNet: the positive linear satisfiability neural networks
Runzhong Wang, Yunhao Zhang, Ziao Guo, Tianyi Chen, Xiaokang Yang, Junchi Yan.
-
Learning generative RNN-ODE for collaborative time-series and event sequence forecasting
Longyuan Li, Junchi Yan, Yunhao Zhang, Jihai Zhang, Jie Bao, Yaohui Jin, Xiaokang Yang
-
Synergetic learning of heterogeneous temporal sequences for multi-horizon probabilistic forecasting
Longyuan Li, Jihai Zhang, Junchi Yan, Yaohui Jin, Yunhao Zhang, Yanjie Duan, Guangjian Tian.
🎖 Honors and Awards
- 2023 Graduate Fellowship of Yang Yuanqing Education Fund (杨元庆教育基金优秀硕士奖学金, 3 in CS Department)
- 2022 National Scholarship (国家奖学金)
- 2021 Undergraduate Honors Scholarship of Yang Yuanqing Education Fund (杨元庆教育基金优秀本科生卓越奖学金, 3 in CS Department)
- 2020 Shanghai Scholarship (上海市奖学金)
- 2020 Meritorious Winner of Mathematical Contest in Modeling (美国大学生数学建模竞赛一等奖, approximately top 8% of teams)
📖 Educations
- 2021.09 - Current, PhD of Comupter Science and Technology, Shanghai Jiao Tong University (SJTU, 上海交通大学)
- 2017.09 - 2021.06, Bachelor of Comupter Science and Technology, Shanghai Jiao Tong University (SJTU, 上海交通大学)
- 2014.09 - 2017.06, Xi’an Gaoxin No.1 High School (西安高新第一中学)
💻 Internships
- 2021.06 - 2021.09, Alibaba DAMO Academy, mentor: Dr. Qingsong Wen (文青松).
- 2025.07 - 2026.01, Huawei Noah’s Ark Laboratory, mentor: Dr. Lujia Pan (潘璐伽) and Dr. Jiale Zheng (郑嘉乐).