Preprints and Working Papers
- Optimizing probabilistic conformal prediction with vectorized non-conformity scores.
M. Zheng and S. Zhu. - Recurrent neural goodness-of-fit test for time series.
A. Zhang, W. Zhou, L. Xie, and S. Zhu. - Balancing optimality and diversity: Human-centered decision-making through generative curation. [Code]
M. L. Li and S. Zhu. Submitted. - New user event prediction through the lens of causal inference.
H. S. Yuchi, S. Zhu, L. Dong, Y. M. Arisoy, and M. C. Spencer. - TimeAutoDiff: Combining autoencoder and diffusion model for time series tabular data synthesizing.
N. Suh, Y. Yang, D. Hsieh, Q. Luan, S. Xu, S. Zhu, and G. Cheng. - Counterfactual fairness through transforming data orthogonal to bias.
S. Chen and S. Zhu. - Conditional generative representation for black-box optimization with implicit constraints. [Poster]
W. Xing, J. Lee, C. Liu, and S. Zhu.- A short version is accepted as a referred paper for 2024 INFORMS Optimization Society Conference.Â
- A short version is accepted by ICLR 2024 Workshop on Generative Models for Decision Making.Â
- Conditional generative modeling for high-dimensional marked temporal point processes. [Code]
Z. Dong, Z. Fan, and S. Zhu.- A short version is accepted by NeurIPS 2023 Workshop on SyntheticData4ML.Â
- Generalized Hypercube queuing models with overlapping service regions. [Code]
S. Zhu, W. Xing, and Y. Xie. Under Review. - Quantifying grid resilience against extreme weather using large-scale customer power outage data. [Poster] [Code]
S. Zhu, R. Yao, Y. Xie, F. Qiu, Y. Qiu, and X. Wu. In Minor Revision. - Multi-resolution spatio-temporal prediction with application to wind power generation. [Slides] [Poster]
Z. Dong, H. Zhang, S. Zhu, Y. Xie, and P. Van Hentenryck. Submitted.Â- Finalist of 2023 QSR Best Student Paper Competition, Awarded to Student Author (Zheng Dong).Â
- Finalist of Best Applied Paper Competition at 2023 INFORMS Workshop on Data Mining and Decision Analytics.
- Best Student Paper Nominee & full presentation at 2022 INFORMS Workshop on Data Science.
Journal
- Non-stationary spatio-temporal point process modeling for high-resolution COVID-19 data. [arXiv] [Poster] [Code]
Z. Dong, S. Zhu, Y. Xie, J. Mateu, and F. J. RodrÃguez-Cortés.
Journal of the Royal Statistical Society, Series C. Vol. 72, Issue 2, Pages 368–386, May 2023.- Winner, 2022 INFORMS Data Mining Best Applied Paper.
- Runner up, 2022 INFORMS Poster Competition.
- Accepted for presentation at 2022 INFORMS Workshop on Data Science.
- Sequential adversarial anomaly detection for one-class event data. [arXiv] [Slides] [Code]
S. Zhu, H. S. Yuchi, M. Zhang, and Y. Xie.
INFORMS Journal on Data Science. Vol 2, No. 1, Pages 45-59, Mar 2023.- Runner up, Best Applied Paper Competition at 2020 INFORMS Workshop on Data Mining and Decision Analytics.
- A short version is accepted for presentation by ICASSP 2021.
- Data-driven optimization for Atlanta police zone design. [arXiv] [Slides] [Code]
S. Zhu, H. Wang, and Y. Xie.
INFORMS Journal on Applied Analytics (formerly Interfaces). Vol. 52, Issue 5, Pages 412-432, Oct 2022.- Finalist of 2021 INFORMS Wagner Prize.Â
- 2nd place in 2019 INFORMS Doing Good with Good OR Best Paper Competition.
- The proposed zone design was adopted by Atlanta Police Department and implemented in 2019.
- Spatio-temporal-textual point processes for crime linkage detection. [arXiv] [Poster] [Code]
S. Zhu and Y. Xie.
Annals of Applied Statistics. Vol. 16, No. 2, Pages 1151-1170, June 2022.- Selected to be presented in The Best of AOAS Session at JSM 2022 (Only three out of all papers).
- Best Poster Presentation Award at 2018 Forecasting from Complexity Workshop.
- Best Student Poster Award at Georgia Statistics Day 2018.
- The project won the 2018 Smart 50 Award at the Smart Cities Connect Conference & Expo.
- The proposed algorithm was implemented by Atlanta Police Department in their AWARE system in 2018.
- Early detection of COVID-19 hotspots using spatio-temporal data. [arXiv] [Slides] [Poster] [Code]
S. Zhu, A. Bukharin, L. Xie, K. Yamin, S. Yang, P. Keskinocak, and Y. Xie.
IEEE Journal of Selected Topics in Signal Processing. Vol. 16, Issue 2, Page 250-260, Feb 2022.- Finalist of Best Applied Paper Competition at 2021 INFORMS Workshop on Data Mining and Decision Analytics.
- Best Paper Award (Honorable Mention) at ICML Time Series Workshop 2021.Â
- A short version is accepted for presentation and highlighted as contributed talk by ICML Time Series Workshop 2021.
- Excellent Poster Award at Georgia Statistics Day 2021.
- Imitation learning of neural spatio-temporal point processes. [arXiv] [Slides] [Poster] [Code]
S. Zhu, S. Li, Z. Peng, and Y. Xie.
IEEE Transactions on Knowledge and Data Engineering. Vol. 34, Issue 11, Nov 2022.- A short version is accepted for presentation by NeurIPS AI for Earth Sciences Workshop 2020.Â
- Spatio-temporal point processes with attention for traffic congestion event modeling. [arXiv] [Code]
S. Zhu, R. Ding, M. Zhang, P. Van Hentenryck, and Y. Xie.
IEEE Transactions on Intelligent Transportation Systems. Vol. 23, Issue 7, July 2022. - In Situ Transmission Electron Microscopy: Signal processing challenges and example. [arXiv]
J. Kacher, Y. Xie, S. P. Voigt, S. Zhu, H. S. Yuchi, J. Key, and S. R. Kalidindi.
IEEE Signal Processing Magazine (Survey paper). Vol. 39, Issue 1, Jan 2022. - High-resolution spatio-temporal model for county-level COVID-19 activity in the U.S. [arXiv] [Code]
S. Zhu, A. Bukharin, L. Xie, M. Santillana, S. Yang, and Y. Xie.
ACM Transactions on Management Information Systems. Vol. 12, Issue 4, Article 33, Dec 2021. - Investigating local oxidation processes in Fe thin films in a water vapor environment by in situ liquid cell TEM.
J. Key, S. Zhu, C. M. Rouleauc, R. R. Unocic, Y. Xie, and J. Kacher.
Ultramicroscopy. Vol. 209, Feb 2020.
Conference
- Counterfactual generative models for time-varying treatments. [arXiv] [Slides] [Poster] [Code]
S. Wu, W. Zhou, M. Chen, and S. Zhu.
ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), 2024.- Spotlight, Deep Generative Models for Health (DGM4H) Workshop at NeurIPS 2023 (Five out of 44 accepted papers).Â
- Winner, 2023 YinzOR Poster Competition, Awarded to Student Author (Wenbin Zhou).
- A short version is accepted by Causal Representation Learning Workshop at NeurIPS 2023.Â
- Distributionally robust weighted k-nearest neighbors. [arXiv] [Poster] [Code] [News]
S. Zhu, L. Xie, M. Zhang, R. Gao, and Y. Xie.
Annual Conference on Neural Information Processing Systems (NeurIPS), 2022. - Neural spectral marked point processes. [arXiv] [Slides] [Poster] [Code] [News]
S. Zhu, H. Wang, Z. Dong, X. Cheng, and Y. Xie.
International Conference on Learning Representations (ICLR), 2022. - Sequential adversarial anomaly detection with deep Fourier kernel. [Slides]
S. Zhu, H. S. Yuchi, M. Zhang, and Y. Xie.
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021. - Deep Fourier kernel for self-attentive point processes. [arXiv] [Slides]
S. Zhu, M. Zhang, R. Ding, and Y. Xie.
International Conference on Artificial Intelligence and Statistics (AISTATS), 2021.- Oral presentation, acceptance rate: 48/1527 = 3.1%.
- Goodness-of-fit test for self-exciting processes. [arXiv]
S. Wei, S. Zhu, M. Zhang, and Y. Xie.
International Conference on Artificial Intelligence and Statistics (AISTATS), 2021. - Adversarial anomaly detection for marked spatio-temporal streaming data. [Slides] [Poster] [Code]
S. Zhu, H. S. Yuchi, and Y. Xie.
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2020.- Best Student Poster Award (Honorable Mention) at Georgia Statistics Day 2019.
- Crime event embedding with unsupervised feature selection. [arXiv] [Poster] [Code]
S. Zhu and Y. Xie.
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2019. - Learning temporal point processes via reinforcement learning. [arXiv] [Code]
S. Li, S. Xiao, S. Zhu, N. Du, Y. Xie, and L. Song.
Annual Conference on Neural Information Processing Systems (NeurIPS), 2018.- Spotlight, acceptance rate: 168/4856 = 3.5%.
- Crime incidents embedding using restricted Boltzmann machines. [arXiv] [Code]
S. Zhu and Y. Xie.
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2018. - Sequential adaptive detection for in-situ transmission electron microscopy (TEM). [arXiv] [Poster]
Y. Cao, S. Zhu, Y. Xie, J. Key, J. Kacher, R. R. Unocic, and C. M. Rouleau.
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2018.
Workshop
- Assessing electricity service unfairness with transfer counterfactual learning.
S. Wei, X. Kong, A. Santos Xavier, S. Zhu, Y. Xie, and F. Qiu.
Causal Representation Learning Workshop at NeurIPS 2023. - Uncertainty-aware robust learning on noisy graphs. [Poster] [Code]
S. Chen, K. Ding, and S. Zhu.
New Frontiers in Graph Learning Workshop at NeurIPS 2023. - Data-driven optimization for police beat design in South Fulton, Georgia. [arXiv] [Slides]
S. Zhu, A. Bukharin, L. Lu, H. Wang, and Y. Xie.
Data Science for Social Good at KDD 2021.Â- The proposed beat design was adopted by South Fulton Police Department and implemented in 2020.
Patent
Y. Xie and S. Zhu. U.S. Patent No. US20190318223A1.
Thesis
- Statistical Learning and Decision Making for Spatio-Temporal Data.Â
Ph.D. dissertation, Georgia Institute of Technology. April 2022.Â
News Coverage
- Data-driven policing
- Atlanta police zone design
- Police beat reconfiguration for City of South Fulton, GA
- Fox 5: Community policing: City of South Fulton gets help from Georgia Tech to shape beats
- City of South Fulton: Georgia Tech police beat design will help cut response times
- Saporta Report: Community policing: City of South Fulton gets help from Georgia Tech to shape beats
- South Fulton police adds cruisers as new beat design puts more officers on the street
- Ga. PD adds 41 vehicles to support new beat design
- COVID-19 Modeling