Haoyi Zhou / 周号益

I am Haoyi Zhou (周号益), currently an associate professor in the College of Software at Beihang University, also with Beijing Advanced Innovation Center for Big Data and Brain Computing (BDBC) in the ACT lab research unit.

My current research interests are in artificial intelligence and machine learning research, with a focus on the application of time-series forecasting, task decision-making, AI safety, AI for Science and other technologies in the industrial field and medical care. We forge the BUAA-SMART group. If you are eager for the similar research vision, please feel free to reach me.

My group has several fully funded PhD positions starting in Fall 2025. Anyone passionate about AI and AI4S research with a focus on SEQUENCES is welcome to apply and reach out. Please send me your CV and a cover letter describing your preferred research topics.

Haoyi Zhou

haoyi [at] buaa.edu.cn

Room G503
New Main Building

Beihang University
37 Xueyuan Road
Haidian District
Beijing, 100191
P.R. China

NEWs

[03/2025] Nominated "2024 Most Cited Chinese Researchers" for exceptional research performance (Scopus)

[12/2024] Awarded "First-Class Natural Science Awards" at the Science and Technology Award of the CIE

[08/2024] We won the "Annual Outstanding Achievement" at the first ZCZ-award

[05/2024] We pre-released the OmniArch-α foundation model at Huawei KADC

[04/2024] We co-held "Competitions: Rapid Aerodynamic Drag Prediction for Arbitrary 3D Vehicle" at IJCAI 2024

[01/2024] We won the national silver award at the 9th China 'Internet+' Innovation and Entrepreneurship Competition

[10/2023] We co-held "Neural symbol computing seminar" and "AI4SC seminar" at CNCC 2023

[09/2023] Nominated "The 9th Young Elite Scientists Sponsorship Program" at CAST recommending by CAAI

[09/2023] We won the second place at the L2RPN Delft 2023 Competition

[06/2023] We won the championship at the China Open Source Innovation Competition

[03/2023] Awarded "Outstanding Doctorate Dissertation" at the CAAI Wenjun Wu Awards

[11/2022] Nominated "CCF Youth Talent Development Program" at the CCF Awards

[08/2022] Awarded "Youth Outstanding Paper Nomination" at the WAIC2022

[05/2022] Our AIQoSer paper has been awarded IWQoS'22 Best Paper : Official

[05/2022] Awarded "Most Popular Speaker of the Year" at the Techbeat.net

[01/2022] We won the championship at the 2nd Cloud Native Programming Challenge Final 阿里云

[08/2021] We won the first prize at the 15th National/University Computer Game Championship

[07/2021] We won the first prize at the 7th China "Internet+" Innovation and Entrepreneurship Competition (Beijing)

[07/2021] Awarded "AI Rising Star" at the World Artificial Intelligence Conference (WAIC)

[06/2021] Awarded "Outstanding Graduate Student" by Beijing Municipal Commission of Education

[06/2021] Awarded "Top-10 Doctoral Students" at Beihang University

[05/2021] Successfully defended my Ph.D. thesis: Generalizable Sequential Data Modeling Method

[02/2021] Nominated "Field Supernova" at AMiner & BAAI

[02/2021] Our Informer paper has been awarded AAAI'21 Best Paper: Official, Beihang, Rutgers

[12/2020] Awarded "Haiying Star" by Haidian District People's Government of Beijing Municipality

[12/2020] We won the first prize at National Industrial Internet APP and Information Consumption Race

[10/2020] Nominated "Outstanding Contribution for Defending Coronavirus" at SCSE of Beihang University

Publications

Towards Objective Fine-tuning: How LLMs' Prior Knowledge Causes Potential Poor Calibration?, ACL 2025

Ziming Wang, Zeyu Shi, Haoyi Zhou, Shiqi Gao, Qingyun Sun, Jianxin Li

[Paper] [OpenReview] [Code] [BibTex]

OmniArch: Building the Foundation Model for Scientific Computing, ICML 2025

Tianyu Chen, Haoyi Zhou, Ying Li, Hao Wang, Chonghan Gao, Rongye Shi, Shanghang Zhang, Jianxin Li

[Paper] [arXiv] [Code] [BibTex]

PINNsAgent: Automated PDE Surrogation with Large Language Models, ICML 2025

Qingpo Wu, Chonghan Gao, Tianyu Chen, Yihang Huang, Yuekai Zhang, Jianing Wang, Jianxin Li, Haoyi Zhou, Shanghang Zhang

[Paper][Code][BibTex]

FreqMoE: Dynamic Frequency Enhancement for Neural PDE Solvers, IJCAI 2025

Tianyu Chen, Haoyi Zhou, Ying Li, Hao Wang, Zhenzhe Zhang, Tianchen Zhu, Shanghang Zhang, Jianxin Li

[Paper] [Code] [BibTex]

Galaxy Walker: Geometry-aware VLMs For Galaxy-scale Understanding, CVPR 2025

Tianyu Chen, Xingcheng Fu, Yisen Gao, Haodong Qian, Yuecen Wei, Kun Yan, Haoyi Zhou, Jianxin Li

[Paper] [arXiv] [Code][BibTex]

FLUE: Streamlined Uncertainty Estimation for Large Language Models, AAAI 2025

Shiqi Gao, Tianxiang Gong, Haoyi Zhou, Zijie Lin, Runhua Xu, Jianxin Li

[Paper] [Code] [BibTex]

MalDetectFormer: Leveraging Sparse SpatioTemporal Information for Effective Malicious Traffic Detection, AAAI 2025

Shuai Zhang, Yu Fan, Haoyi Zhou, Bo Li

[Paper] [Code] [BibTex]

QUEST: Quadruple Multimodal Contrastive Learning with Constraints and Self-Penalization, NeurIPS 2024

Qi Song, Tianxiang Gong, Shiqi Gao, Haoyi Zhou, Jianxin Li

[Paper] [Poster] [Code] [BibTex]

From Text to Trajectory: Exploring Unified Complex Constraint Representation and Decomposition in Safe Reinforcement Learning, NeurIPS 2024

Pusen Dong, Tianchen Zhu, Yue Qiu, Haoyi Zhou, Jianxin Li

[Paper] [arXiv] [Poster] [Code] [BibTex]

ArcEDB: An Arbitrary-Precision Encrypted Database via (Amortized) Modular Homomorphic Encryption, ACM CCS 2024

Zhou Zhang, Song Bian, Zian Zhao, Ran Mao, Haoyi Zhou, Jiafeng Hua, Yier Jin, Zhenyu Guan

[Paper] [Code] [BibTex]

PE-Attack: On the Universal Positional Embedding Vulnerability in Transformer-based Models, IEEE TIFS 2024

Shiqi Gao, Haoyi Zhou, Tianyu Chen, Mingrui He, Runhua Xu, Jianxin Li

[Paper] [Code] [BibTex]

Decoding Global Preferences: Temporal and Cooperative Dependency Modeling in Multi-agent Preference-based Reinforcement Learning, AAAI 2024

Tianchen Zhu, Yue Qiu, Haoyi Zhou, Jianxin Li

[Paper] [PDF] [Code] [BibTex]

Self-Paced Unified Representation Learning for Hierarchical Multi-Label Classification, AAAI 2024

Zixuan Yuan, Hao Liu, Haoyi Zhou, Denghui Zhang, Xiao Zhang, Hao Wang, Hui Xiong

[Paper] [PDF] [Code] [BibTex]

PhoGAD: Graph-Based Anomaly Behavior Detection with Persistent Homology Optimization, WSDM 2024

Ziqi Yuan, Haoyi Zhou, Tianyu Chen, Jianxin Li

[Paper] [arXiv] [Code] [BibTex]

MultiNetAD: Multiplex Network-Based Anomaly Access Detection Featuring Semantic Hierarchies, SDM 2024

Ziqi Yuan, Qingyun Sun, Haoyi Zhou, Zukun Zhu, Jianxin Li

[Paper] [Code] [BibTex]

LEVER: Online Adaptive Sequence Learning Framework for High-Frequency Trading, TKDE 2023

Zixuan Yuan, Junming Liu, Haoyi Zhou, Denghui Zhang, Hao Liu, Nengjun Zhu, Hui Xiong

[Paper] [Code] [BibTex]

Towards Long-delayed Sparsity: Learning a Better Transformer through Reward Redistribution, IJCAI 2023

Tianchen Zhu, Yue Qiu, Haoyi Zhou, Jianxin Li

[Paper] [Code] [BibTex]

Expanding the Prediction Capacity in Long Sequence Time-Series Forecasting, AIJ 2023

Haoyi Zhou, Jianxin Li, Shanghang Zhang, Shuai Zhang, Mengyi Yan, Hui Xiong

[Paper] [Code] [BibTex]

BadRes: Reveal the Backdoors through Residual Connection, ICASSP 2023

Mingrui He, Tianyu Chen, Haoyi Zhou, Shanghang Zhang, Jianxin Li

[Paper] [arXiv] [Code] [BibTex]

CLDG: Contrastive Learning on Dynamic Graphs, ICDE 2023

Yiming Xu, Bin Shi, Teng Ma, Bo Dong, Haoyi Zhou, Qinghua Zheng

[Paper] [Code] [BibTex]

Jump Self-attention: Capturing High-order Statistics in Transformers, NeurIPS 2022

Haoyi Zhou, Siyang Xiao, Shanghang Zhang, Jieqi Peng, Shuai Zhang, Jianxin Li

[Paper] [Openreview] [Slides] [Poster] [Code] [BibTex]

AutoST: Towards the Universal Modeling of Spatio-temporal Sequences, NeurIPS 2022

Jianxin Li, Shuai Zhang, Hui Xiong, Haoyi Zhou

[Paper] [Openreview] [Spotlight Talk] [Poster] [Code] [BibTex]

AIQoSer: Building the efficient Inference-QoS for AI Services, IEEE IWQoS 2022

Jianxin Li, Tianchen Zhu, Haoyi Zhou, Qingyun Sun, Chunyang Jiang, Shuai Zhang, Chunming Hu

[Paper] [Code] [BibTex]

THE-X: Privacy-Preserving Transformer Inference with Homomorphic Encryption, ACL Findings 2022

Tianyu Chen, Hangbo Bao, Shaohan Huang, Li Dong, Binxing Jiao, Daxin Jiang, Haoyi Zhou, Jianxin Li, Furu Wei

[Paper] [Code] [BibTex]

Learning Music Sequence Representation from Text Supervision, ICASSP 2022

Tianyu Chen, Yuan Xie, Shuai Zhang, Shaohan Huang, Haoyi Zhou, Jianxin Li

[Paper] [Code] [BibTex]

MTTrans: Cross-Domain Object Detection with Mean Teacher Transformer, ECCV 2022

Jinze Yu, Jiaming Liu, Xiaobao Wei, Haoyi Zhou, Yohei Nakata, Denis Gudovskiy, Tomoyuki Okuno, Jianxin Li, Kurt Keutzer, Shanghang Zhang

[Paper] [arXiv] [BibTex]

Informer: Beyond Efficient Transformer for Long Sequence Timer-series Forecasting, AAAI'2021

Haoyi Zhou, Shanghang Zhang, Jieqi Peng, Shuai Zhang, Jianxin Li, Hui Xiong, Wancai Zhang

[Paper] [arXiv] [Poster] [Code] [Dataset] [Video] [Techbeat+Slides] [AiTime] [PaperWeekly] [BibTex]

Triplet Attention: Rethinking the Similarity in Transformers, KDD'2021

Haoyi Zhou, Shanghang Zhang, Jieqi Peng, Shuai Zhang, Jianxin Li

[Paper] [Slides] [Code] [ACM Video] [Techbeat+Slides] [BibTex]

POLLA: Enhancing the Local Structure Awareness in Long Sequence Spatial-temporal Modeling, ACM TIST 2021

Haoyi Zhou, Hao Peng, Jieqi Peng, Shuai Zhang, Jianxin Li

[Paper] [Code] [Slides] [BibTex]

MERITS: Medication Recommendation for Chronic Disease with Irregular Time-Series, ICDM'2021

Shuai Zhang, Jianxin Li, Haoyi Zhou, Qishan Zhu, Shanghang Zhang, Danding Wang

[Paper] [Code] [BibTex]

MtCut: A Multi-Task Framework for Ranked List Truncation, WSDM'2021

Dong Wang, Jianxin Li, Tianchen Zhu, Haoyi Zhou, Qishan Zhu, Yuxin Wen, Hongming Piao

[Paper] [Code] [BibTex]

RoSearch: Search for Robust Student Architectures When Distilling Pre-trained Language Models, arXiv 2021

Xin Guo, Jianlei Yang, Haoyi Zhou, Xucheng Ye, Jianxin Li

[Paper] [Slides] [BibTex]

Osprey: A Fast and Accurate Patch Presence Test Framework for Binaries, Computer Communications 2021

Peiyuan Sun, Qiben Yan, Haoyi Zhou*, Jianxin Li

[Paper] [BibTex]

Differentially-private Federated Neural Architecture Search, arXiv 2020

Ishika Singh, Haoyi Zhou, Kunlin Yang, Meng Ding, Bill Lin, Pengtao Xie

[Paper] [Code] [BibTex]

A Bisubmodular Approach to Event Detection and Prediction in Multivariate Social Graphs, IEEE TCSS 2020

Shuai Zhang, Haoyi Zhou, Feng Chen, Jianxin Li

[Paper] [BibTex]

A Time-Sensitive Hybrid Learning Model for Patient Subgrouping, IJCNN'2018

Yingchun Zhang, Haoyi Zhou, Jianxin Li, Wanlu Sun, Yahong Chen

[Paper] [BibTex]

Improving the Generalization Performance of Multi-class SVM via Angular Regularization, IJCAI'2017

Jianxin Li, Haoyi Zhou, Pengtao Xie, Yingchun Zhang

[Paper] [Poster] [Appendix] [BibTex]

Stacked Kernel Network, arXiv 2017

Shuai Zhang, Jianxin Li, Pengtao Xie, Yingchun Zhang, Minglai Shao, Haoyi Zhou, Mengyi Yan

[Paper] [BibTex]

A Large Margin Learning Method for Matching Images of Natural Objects with Different Dimensions, CVPREI 2016

Haoyi Zhou, Jun Zhou, Haichuan Yang, Cheng Yan, Xiao Bai and Yunlu Liu

[Paper] [BibTex]

Teaching

Database Management Technology, Undergraduate Course, 2023/2024 Spring

Big Data and Cloud Computing, Undergraduate Course, 2023 Spring

Programming and Algorithms, Graduate Course, 2022/23 Fall

Practical Programming, Undergraduate Course, 2023 Summer

Frontier Technology of AI Open Source Computing System, Graduate Course, 2022-24 Spring

Industrial Internet and Big Data, Graduate Course, 2022-24 Spring

Projects

NSFC, PI, Research on Long Sequence Forecasting for Industrial Big Data, 2022

Big data technology is widely applied in industrial digital transformation, and the growth in both the length and scale of sequence data raises the need for long sequence analysis. However, the traditional methods have severe long-term prediction failure and other problems. Based on the leading achievements in our proposed self-attention mechanism, this project plans to focus on the long sequence problems.

[Homepage][News][Sponsor]

NKRDP, PI, Sequence Data Processing Technology, 2022

Preserved.

[Homepage][News][Sponsor]

State-Grid, Co-PI, Supply Chain Construction and Value Mining, 2023

This research proposes a technical system for building and exploring the value of the green and intelligent supply chain. It systematically guides the transformation of enterprises along the chain toward a green and intelligent direction, providing vital support for the high-quality development of the electricity company.

[Homepage][News][Dataset][Sponsor]

Huawei, PI, Kit for Long Sequence Intelligent Computing, 2022

Preserved.

[Homepage][News][Dataset][Sponsor]

Huawei, Co-PI, AI Scientific Computing Software, 2022

Preserved.

[Homepage][News][Dataset][Sponsor]

CAAI-Huawei MindSpore Open Fund, PI, Research on Long Sequence Decision, 2022

This project focuses on the core scientific problem of "Long Sequence Prediction and Decision-making." With original innovation in self-attention mechanisms as its driving force, it distinguishes itself from the traditional approach of decomposing and solving local problems. By unlocking the modeling capabilities, activating collaborative abilities, and strengthening compositional skills, this project not only enhances the fundamental understanding of long sequence prediction but also yields practical insights through feedback and adjustments in real industrial application scenarios.

[Homepage][News][Dataset][Sponsor]

Education

Visiting Scholar in Computer Science

Shenzhen Institute of Computing Sciences (Dec 2021 – June 2022, Shenzhen, China), advisor: Prof. Wenfei Fan

PhD in Computer Software and Theory

Beihang University (Sep 2015 – June 2021, Beijing, China), advisor: Prof. Jianxin Li
Thesis: Generalizable Sequential Data Modeling Method [Link]

Visiting Scholar in Computer Science

Rutgers University (Aug 2019 – Sep 2020, Newark, US), advisor: Prof. Xiong Hui

MSc in Computer Science and Technology

Beihang University (Sep 2013 – Aug 2015, Beijing, China), advisor: Prof. Xiao Bai
Thesis: Graphical Matching and Retrieval of Images [Transfered]

BSc in Aircraft Design and Engineering

Beihang University (Sep 2009 – July 2013, Beijing, China)
Thesis: Numerical Algorithms for Solving Two-point Boundary Problems

Talks

Invited Talk: The OmniArch Equation Solving model Based on MindSpore Framework, Ascend Developer Summit, May 11, 2024

Invited Talk: AI for scientific computing: from the sequential architecture, The 7th Global Intelligent Industry Conference, Mar 31, 2024

Invited Talk: Building AI for scientific computing platform, The 6th Xiuhu Conference, Oct 20, 2023

Spotlight Presentation: AutoST: Towards the Universal Modeling of Spatio-temporal Sequences, NeurIPS (online), Dec 7, 2022

Oral Presentation: Triplet Attention: Rethinking the similarity in Transformers, KDD (online), Aug 17, 2021

Invited Talk: Exploring the long sequence time-series forecasting, Nvidia Symposium, June 30, 2021

Invited Talk: 长序列预测向左，计算复杂度向右, Tianjin University (北洋智算论坛), June 6, 2021

Oral Presentation: Informer: Beyond efficient transformer for long sequence time-series forecasting, AAAI (online), Feb 9, 2021

Invited Talk: The Informer in Long Sequence Time-series Forecasting, BAAI (智源预讲会), Dec 9, 2020

Oral Talk: The Generalization Performance in Machine learning, IWCST (Beijing), Nov 12, 2017

Oral Presentation: Improving the Generalization Performance of Multi-class SVM via Angular Regularization, IJCAI (Melbourne), Aug 23, 2017

Selected Media

LibHunt. Top 22 Python time-series Projects (Ranked 4th), June 1, 2021

Topbots. Top Research Papers With Business Applications, April 12, 2021

TowardsDataScience. Adding the Informer Model to Flow Forecast, April 8, 2021

Qita. 時系列データTransformer論文３選を読んでみた, March 23, 2021

Synced & Medium. AAAI 2021 Best Papers Announced, Feb 4, 2021

将门创投. 从“鹦鹉”到“乌鸦”，AI的本质是探索通用智能的可能性, May 7, 2022

文汇报. 有「最好土壤」的上海，集聚AI「最牛尖子生」, July 12, 2021

澎湃新闻. 这场AI盛会中，后浪与前浪都在讨论什么？, July 12, 2021

解放日报. 数字化时代AI冲浪者，追求产业变革最优解, July 10, 2021

人民日报. 全球AI青年人才论坛：改变世界从改变产业做起, July 9, 2021

中国新闻网. AI青年科学家联盟“A班”朋友圈扩容：中国AI大有可为, July 9, 2021

科技日报. 让数据支撑抗议科学决策——记北航新冠疫情大数据分析团队, June 6, 2020

Academic Service

Journal Reviewer

Nature Com.Eng, Science China, ACM Computing Survey
IEEE TPAMI, IEEE TKDE, Machine Learning, AIJ
IEEE TNNLS, ACM TIST, IEEE TCSS
Neurocomputing, KBS, Neural Networks, IEEE RA-L

Conference Reviewer

NIPS 2021-26, ICML 2021-26, ICCV 2022-24, ECCV 2022-23
KDD 2021-26, ICDM 2021, CIKM 2020-21
AAAI 2020-26, IJCAI 2020-26, ICLR 2022-25, AISTATS 2025

Others

The cheif commitee of IJCAI-24 competition, 2024
The youth assistant editor of Frontiers of Computer Science, 2022-now
The member of Beijing Youth Internet Association, 2021-now
The chairperson of CCF Beihang Student Branch, 2014-2015 [link]