Donglin  Zhan

RECENT RESEARCH INTERESTS

(1) Deep learning for alpha generation and time-series modeling in quantitative trading. I focus less on novel architectures or training paradigms for their own sake, and more on what actually makes a model survive in the extremely low signal-to-noise regimes of real markets — the full loop of training, loss design, calibration, and the inference infrastructure around it. I'm especially interested in where deep learning can complement, rather than replace, traditional cross-sectional analysis and statistical/optimization methods. On the side, I also explore on-chain alternative data and prediction markets, looking for cases where price diverges from fair value under social-momentum dynamics.

(2) Foundation models for climate and geophysical science. How to inject appropriate physical priors during (post-)training, so that models deliver not just accurate predictions but explainability and well-calibrated confidence.

SOME RECENT RESEARCH 

Leonardo F. Toso, Davit Shadunts, Yunyang Lu, Nihal Sharma, Donglin Zhan, Nam H Nguyen, James Anderson   Learning Invariant Visual Representations for Planning with Joint-Embedding Predictive World Models [paper]

Donglin Zhan, Haoting Zhang, Rhonda Righter, Zeyu Zheng, James Anderson   Collaborative Bayesian Optimization with Wasserstein Barycenters IEEE Conference on Decision and Control 2025 (CDC) [paper]

Donglin Zhan, Leonardo F. Toso, James Anderson   Coreset-Based Task Selection for Sample-Efficient Meta-Reinforcement Learning  IEEE Conference on Decision and Control 2025 (CDC) [paper]

Haoting Zhang*, Donglin Zhan*, James Anderson, Rhonda Righter, Zeyu Zheng   Clustering then Estimation of the Spatio-Temporal Self-Exciting Processes, INFORMS Journal on Computing [paper]

Leonardo F. Toso, Donglin Zhan, James Anderson, Han Wang Meta-Learning Linear Quadratic Regulators: A Policy Gradient MAML Approach for the Model-free LQR, Learning for Dynamics and Control Conference 2024 (L4DC) (Best Paper Award) [paper]

Donglin Zhan, James Anderson  Data-Efficient and Robust Task Selection for Meta-Learning,  IEEE/CVF Conference on Computer Vision and Pattern Recognition 2024 (CVPR) Workshop [paper] 

Haoting Zhang, Haoxian Chen, Donglin Zhan, Hanyang Zhao, Henry Lam, Wenpin Tang, David Yao, Zeyu Zheng   SOCRATES: Simulation Optimization with Correlated Replicas and Adaptive Trajectory Evaluations, Short Version on NeurIPS 2025 MLxOR Workshop [paper]

Haoting Zhang*, Donglin Zhan*, Yunduan Lin, Jinghai He, Qing Zhu, Max Zuo-Jun Shen, Zeyu Zheng  Daily Physical Activity Monitoring—Adaptive Learning from Multi-source Motion Sensor Data, Conference on Health, Inference, and Learning (CHIL) 2024  [paper]

Donglin Zhan, Yusheng Dai, Yiwei Dong, Jinghai He, Zhenyi Wang, James Anderson  Meta-Adaptive Stock Movement Prediction with Two-Stage Representation Learning, Workshop on Distribution Shift, Neural Information Processing System 2022 (NeurIPS), SIAM Conference on Data Mining 2024 (SDM)  [paper]

Yunkai Zhang, Donglin Zhan, Haoting Zhang, Max Zuo-Jun Shen, Zeyu Zheng, Qing Zhu  Does Attention Help Wildfire Prediction,  INFORMS Data Mining and Data Analysis Workshop 2024 [paper]

Zhenyi Wang, Li Shen, Donglin Zhan, Qiuling Suo, Yanjun Zhu, Tiehang Duan, Mingchen Gao  MetaMix: Towards Corruption-Robust Continual Learning with Temporally Self-Adaptation Data Transformation, IEEE/CVF Conference on Computer Vision and Pattern Recognition 2023 (CVPR)  [paper] 

Zhenyi Wang, Li Shen, Tongliang Liu, Tiehang Duan, Yanjun Zhu, Donglin Zhan, David Doermann, Mingchen Gao  Defending against Data-Free Model Extraction by Distributionally Robust Defensive Training, Neural Information Processing Systems 2023 (NeurIPS)  [paper] 

Lizi Zhang*, Donglin Zhan*   Stock Movement Prediction via Multi-source Transfer with Covariate Shift Detection, Workshop on ML in Finance, International Conference on Knowledge Discovery and Data Mining 2022 (SIGKDD)  [paper]  

Zhenyi Wang, Li Shen, Tiehang Duan, Donglin Zhan, Le Fang, Mingchen Gao  Learning to Learn and Remember Super Long Multi-Domain Task Sequence, IEEE/CVF Conference on Computer Vision and Pattern Recognition 2022 (CVPR) (Oral)  [paper] 

Zhenyi Wang, Li Shen, Le Fang, Qiuling Suo, Donglin Zhan, Tiehang Duan, Mingchen Gao  Meta-Learning with Less Forgetting on Large-Scale Non-Stationary Task Distributions, European Conference on Computer Vision 2022 (ECCV)  [paper] 

Haoting Zhang, Jinghai He, Donglin Zhan, Zeyu Zheng  Neural Network-Assisted Simulation Optimization with Covariates,  Winter Simulation Conference 2021 (WSC)  [paper]

......

PEER REVIEWER 

Conference: NeurIPS, ICLR, ICML, CVPR, AISTATS, VLDB, ICDE, ECML-PKDD, ICASSP, L4DC

Journal: Neurocomputing, Pattern Recognition, IEEE Transactions on Vehicular Technology, IEEE Transactions on Automatic Control