Peizhong's Homepage

About Me

Hello, I'm Peizhong Ju, an Assistant Professor in the Department of Computer Science of University of Kentucky. I earned my Ph.D. degree from Purdue University in 2021 and my B.S. degree from Peking University in 2016.

My research area includes machine learning, smart grid, optimization, and wireless communication. My recent research focuses on theories for explaining the performance of machine learning models. Before that, I worked on power systems and wireless communication. Generally speaking, the goal of my research is to use rigorous mathematical analysis (including probability theory, optimization, game theory, and random matrix theory) to understand the fundamental limits of a complex system under uncertainty and/or disturbance.

You can contact me at peizhong.ju@uky.edu

My CV (as a pdf file) is here (Last updated: Mar 2, 2025).

Check out my Google Scholar profile for a list of my publications.

News

On Feb 28, 2025, I presented our PSMGD work and chaired an oral session at AAAI'25 at Pennsylvania Convention Center, Philadelphia, Pennsylvania.
On Feb 11, 2025, our paper "How to Find the Exact Pareto Front for Multi-Objective MDPs?" is selected to be presented as a Spotlight (5.1% of the submitted papers) by ICLR'25.
On Jan 22, 2025, three of our papers were accepted by ICLR'25: two focused on diffusion models and one exploring multi-objective MDPs.
On Jan 15-16, 2025, I attended the NeTS Early Career Workshop 2025 at NSF in Alexandria, VA
On Dec 9, 2024, our research paper on Fast Multi-Objective Optimization was accepted to AAAI'25.
I joined the Department of Computer Science of University of Kentucky in August, 2024. Photo
On Jul 28, 2024, two of our papers were accepted by MobiHoc'24. One is about Federated Learning, and another is on network-edge classification.
On Jan 16, 2024, two of my papers on Reinforcement Learning were accepted by ICLR'24.
On Jul 27, 2023, I and Sen Lin presented our continual learning work at ICML'23 at the Hawaii Convention Center, Honolulu.
On Jul 20, 2023, The Ohio State University News reported our continual learning research in a press release: "Future AI algorithms have potential to learn like humans, say researchers."

Experience

Assistant Professor

2024 - Present Department of Computer Science at University of Kentucky, Lexington, KY, USA

Postdoctoral Scholar

2021 - 2024 ECE department at The Ohio State University, Columbus, OH, USA NSF AI Institute on Future Edge Networks and Distributed Intelligence (AI-EDGE)

Ph.D. in Electrical and Computer Engineering

2016 - 2021 Purdue University, West Lafayette, IN, USA

B.Sc. in Electrical Engineering

2012 - 2016 Peking University, Beijing, China

Honors & Awards

Best Paper Award, 13^th ACM International Conference on Future Energy Systems (ACM e-Energy 2022)
The Bilsland Dissertation Fellowship, Purdue University, 2021
Best Paper Award Finalist of 9^th ACM International Conference on Future Energy Systems (ACM e-Energy 2018)
Outstanding Graduate Award of EE department of Peking University, 2016 (the only one that year)
Outstanding Graduate Award of Peking University, 2016
IEEE Student Grant, 2015
Research Excellence Award, Peking University, 2015
China National Scholarship, 2015
Samsung Scholarship, 2014
Award of excellence, OpenHW2014 open source hardware and embedded computing contest (awarded by ARM and XILINX), 2014
EMC Scholarship, 2013

Research

Postdoctoral Scholar, the Ohio State University
Supervisor: Prof. Ness B. Shroff.
- Multi-agent Reinforcement Learning, Multi-task supervised learning, Generalization Performance Analysis. (Oct. 2021 - Aug. 2024)
Research Assistant, Purdue University
Advisor: Prof. Xiaojun Lin.
- Overfitting in Machine Learning (Oct. 2019 - Sept. 2021)
- Market Design for Renewable in Power Distribution Level (Aug. 2018 - Oct. 2019)
- Voltage Control in Power Grid with Renewable (Aug. 2016 - Aug. 2018)
Undergraduate Research Assistant, Peking University, China
Advisor: Prof. Xiang Cheng.
- Antenna Grouping for Massive MIMO (Sep. 2015 - June 2016)
- Self-Interference Cancellation for Full-Duplex (Oct. 2013 - Sep. 2015)

Publications

Conference Papers

Yining Li, Peizhong Ju, and Ness B. Shroff, "How to Find the Exact Pareto Front for Multi-Objective MDPs?" 13^th International Conference on Learning Representations (ICLR'25). spotlight arXiv version
```
@article{li2024find,
  title={How to Find the Exact Pareto Front for Multi-Objective MDPs?},
  author={Li, Yining and Ju, Peizhong and Shroff, Ness B},
  journal={arXiv preprint arXiv:2410.15557},
  year={2024}
}
```

Yuchen Liang, Peizhong Ju, Yingbin Liang, and Ness B. Shroff, "Theory on Score-Mismatched Diffusion Models and Zero-Shot Conditional Samplers," 13^th International Conference on Learning Representations (ICLR'25). arXiv version

@article{liang2024theory,
  title={Theory on Score-Mismatched Diffusion Models and Zero-Shot Conditional Samplers},
  author={Liang, Yuchen and Ju, Peizhong and Liang, Yingbin and Shroff, Ness},
  journal={arXiv preprint arXiv:2410.13746},
  year={2024}
}

Yuchen Liang, Peizhong Ju, Yingbin Liang, and Ness B. Shroff, "Broadening Target Distributions for Accelerated Diffusion Models via a Novel Analysis Approach," 13^th International Conference on Learning Representations (ICLR'25). arXiv version

@article{liang2024broadening,
  title={Broadening target distributions for accelerated diffusion models via a novel analysis approach},
  author={Liang, Yuchen and Ju, Peizhong and Liang, Yingbin and Shroff, Ness},
  journal={arXiv preprint arXiv:2402.13901},
  year={2024}
}

Mingjing Xu, Peizhong Ju, Jia Liu, and Haibo Yang, "PSMGD: Periodic Stochastic Multi-Gradient Descent for Fast Multi-Objective Optimization," 39^th AAAI Conference on Artificial Intelligence (AAAI'25). arXiv version

@article{xu2024psmgd,
  title={PSMGD: Periodic Stochastic Multi-Gradient Descent for Fast Multi-Objective Optimization},
  author={Xu, Mingjing and Ju, Peizhong and Liu, Jia and Yang, Haibo},
  journal={arXiv preprint arXiv:2412.10961},
  year={2024}
}

Peizhong Ju, Haibo Yang, Jia Liu, Yingbin Liang, and Ness B. Shroff, "Can We Theoretically Quantify the Impacts of Local Updates on the Generalization Performance of Federated Learning?" 25^th International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing (MobiHoc'24), 2024 . arXiv version

@inproceedings{ju2024can,
  title={Can We Theoretically Quantify the Impacts of Local Updates on the Generalization Performance of Federated Learning?},
  author={Ju, Peizhong and Yang, Haibo and Liu, Jia and Liang, Yingbin and Shroff, Ness},
  booktitle={Proceedings of the Twenty-fifth International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing (MobiHoc'24)},
  pages={141--150},
  year={2024}
}

Chengzhang Li, Peizhong Ju, Atilla Eryilmaz, and Ness B. Shroff, "Efficient Multi-dimensional Compression for Network-edge Classification," 25^th International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing (MobiHoc'24), 2024 . paper

@inproceedings{li2024efficient,
  title={Efficient Multi-dimensional Compression for Network-edge Classification},
  author={Li, Chengzhang and Ju, Peizhong and Eryilmaz, Atilla and Shroff, Ness},
  booktitle={Proceedings of the Twenty-fifth International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing (MobiHoc'24)},
  pages={91--100},
  year={2024}
}

Peizhong Ju, Arnob Ghosh, and Ness B. Shroff, "Achieving Fairness in Multi-Agent MDP Using Reinforcement Learning," 12^th International Conference on Learning Representations (ICLR'24), 2024. arXiv version

@inproceedings{
  ju2024achieving,
  title={Achieving Fairness in Multi-Agent {MDP} Using Reinforcement Learning},
  author={Peizhong Ju and Arnob Ghosh and Ness Shroff},
  booktitle={The Twelfth International Conference on Learning Representations},
  year={2024},
  url={https://openreview.net/forum?id=yoVq2BGQdP}
}

Yining Li, Peizhong Ju, and Ness B. Shroff, "Achieving Sample and Computational Efficient Reinforcement Learning by Action Space Reduction via Grouping," 12^th International Conference on Learning Representations (ICLR'24), 2024. arXiv version

@inproceedings{liachieving,
  title={Achieving Sample and Computational Efficient Reinforcement Learning by Action Space Reduction via Grouping},
  author={Li, Yining and Ju, Peizhong and Shroff, Ness},
  booktitle={The Twelfth International Conference on Learning Representations},
  year={2024},
  url={https://openreview.net/forum?id=MOmqfJovQ6}
}

Sen Lin, Peizhong Ju (co-first author), Yingbin Liang, and Ness B. Shroff, "Theory on Forgetting and Generalization of Continual Learning," 40^th International Conference on Machine Learning (ICML'23), 2023. arXiv version

@inproceedings{lin2023theory,
  title={Theory on forgetting and generalization of continual learning},
  author={Lin, Sen and Ju, Peizhong and Liang, Yingbin and Shroff, Ness},
  booktitle={International Conference on Machine Learning},
  pages={21078--21100},
  year={2023},
  organization={PMLR}
}

Peizhong Ju, Yingbin Liang, and Ness B. Shroff, "Theoretical Characterization of the Generalization Performance of Overfitted Meta-Learning," 11^th International Conference on Learning Representations (ICLR'23), 2023. arXiv version

@inproceedings{jutheoretical,
  title={Theoretical Characterization of the Generalization Performance of Overfitted Meta-Learning},
  author={Ju, Peizhong and Liang, Yingbin and Shroff, Ness},
  booktitle={The Eleventh International Conference on Learning Representations},
  year={2023},
  url={https://openreview.net/forum?id=Jifob4dSh99}
}

Peizhong Ju, Xiaojun Lin, and Ness B. Shroff, "On the Generalization Power of Overfitted Three-Layer Neural Tangent Kernel Model," 36^th Conference on Neural Information Processing Systems (NeurIPS'22), 2022. arXiv version

@article{ju2022generalization,
  title={On the generalization power of the overfitted three-layer neural tangent kernel model},
  author={Ju, Peizhong and Lin, Xiaojun and Shroff, Ness},
  journal={Advances in neural information processing systems},
  volume={35},
  pages={26135--26146},
  year={2022}
}

Peizhong Ju, Xiaojun Lin, and Jianwei Huang, "Distribution-Level Markets under High Renewable Energy Penetration," Proceedings of the 13^th ACM International Conference on Future Energy Systems (e-Energy'22), 2022. Best Paper Award open access

@inproceedings{ju2022distribution,
  title={Distribution-level markets under high renewable energy penetration},
  author={Ju, Peizhong and Lin, Xiaojun and Huang, Jianwei},
  booktitle={Proceedings of the Thirteenth ACM International Conference on Future Energy Systems},
  pages={127--156},
  year={2022}
}

Peizhong Ju, Xiaojun Lin, and Ness B. Shroff, "On the Generalization Power of Overfitted Two-Layer Neural Tangent Kernel Models," Proceedings of the 38^th International Conference on Machine Learning (ICML'21), PMLR 139:5137-5147, 2021. arXiv version

@inproceedings{ju2021generalization,
  title={On the generalization power of overfitted two-layer neural tangent kernel models},
  author={Ju, Peizhong and Lin, Xiaojun and Shroff, Ness},
  booktitle={International Conference on Machine Learning},
  pages={5137--5147},
  year={2021},
  organization={PMLR}
}

Peizhong Ju, Xiaojun Lin, and Jia Liu, "Overfitting Can Be Harmless for Basis Pursuit, But Only to a Degree," accepted as spotlight presentation at 34^th Conference on Neural Information Processing Systems (NeurIPS'20), 2020 (acceptance rate: 20%, spotlight presentations: 3%). spotlight arXiv version
```
@article{ju2020overfitting,
  title={Overfitting can be harmless for basis pursuit, but only to a degree},
  author={Ju, Peizhong and Lin, Xiaojun and Liu, Jia},
  journal={Advances in Neural Information Processing Systems},
  volume={33},
  pages={7956--7967},
  year={2020}
}
```
Peizhong Ju, and Xiaojun Lin, "Adversarial Attacks to Distributed Voltage Control in Power Distribution Networks with DERs,", Proceedings of the 9^th International Conference on Future Energy Systems (e-Energy'18), pp. 291-302, Karlsruhe, Germany, June 12-15, 2018. Best Paper Award Finalist paper
```
@inproceedings{ju2018adversarial,
  title={Adversarial attacks to distributed voltage control in power distribution networks with DERs},
  author={Ju, Peizhong and Lin, Xiaojun},
  booktitle={Proceedings of the Ninth International Conference on Future Energy Systems},
  pages={291--302},
  year={2018}
}
```
Peizhong Ju, Meng Zhang, Xiang Cheng, and Liuqing Yang, "Generalized spatial modulation with transmit antenna grouping for massive MIMO," Proceedings of 2017 IEEE International Conference on Communications (ICC'17), Paris, France, May 21-25, 2017.
Peizhong Ju, Meng Zhang, Xiang Cheng, Cheng-Xiang Wang, and Liuqing Yang, "Generalized spatial modulation with transmit antenna grouping for correlated channels," Proceedings of 2016 IEEE International Conference on Communications (ICC'16), Kuala Lumpur, Malaysia, May 22-27, 2016.
Peizhong Ju, Miaowen Wen, Xiang Cheng, and Liuqing Yang, "An Effective Self-Interference Cancellation Scheme for Spatial Modulated Full Duplex Systems," Proceedings of 2015 IEEE International Conference on Communications (ICC'15), London, UK, June 8-12, 2015.

Journal Papers

Peizhong Ju, Chengzhang Li (co-first author), Yingbin Liang, and Ness B. Shroff (the last two authors represent the entire AI-EDGE faculty team), "AI-EDGE: An NSF AI Institute for Future Edge Networks and Distributed Intelligence." AI Magazine 45.1 (2024): 29-34. open access
Peizhong Ju, Miaowen Wen, Xiang Cheng, and Liuqing Yang, "Achievable-Rate-Enhancing Self-Interference Cancellation for Full-Duplex Communications," IEEE Transactions on Wireless Communications, vol. 17 no. 12, pp. 8473-8484, 2018.
Weilin Qu, Meng Zhang, Xiang Cheng, and Peizhong Ju, "Generalized Spatial Modulation With Transmit Antenna Grouping for Massive MIMO," IEEE Access, vol. 5, pp. 26798-26807, 2017.

Pending Papers

Muhammad Umair Haider, Hammad Rizwan, Hassan Sajjad, Peizhong Ju, and A.B. Siddique, "Neurons Speak in Ranges: Breaking Free from Discrete Neuronal Attribution." arXiv version
Peizhong Ju, Sen Lin, Mark S. Squillante, Yingbin Liang, and Ness B. Shroff, "Generalization Performance of Transfer Learning: Overparameterized and Underparameterized Regimes." arXiv version
Chengzhang Li, Peizhong Ju, Atilla Eryilmaz, and Ness B. Shroff, "Two Levels Are All You Need: Simplifying Data Compression for Timely Edge Classification."

Book Chapters

Peizhong Ju, Xiaojun Lin, and Ness B. Shroff, "On the Generalization Power of Overfitted Two-Layer Neural Tangent Kernel Models," Chapter 3 of Artificial Intelligence for Edge Computing, Springer Nature, August 2023. DOI: 10.1007/978-3-031-40787-1_3

Patents

Xiang Cheng, Peizhong Ju, and Miaowen Wen, "A method of self-interference cancellation scheme for full duplex systems," 2015100550190, State Intellectual Property Office of the People's Republic of China.

Talks & Presentations

Keeping Current (KCS) talks
Organizer: Department of Computer Science, University of Kentucky
Location: Lexington, KY
Date: Feb 26, 2025
Title: "Advancements in Multi-Objective Optimization: Exact Pareto Fronts in MDPs and Efficient Gradient Descent Methods"
UK Applied Math Seminar
Organizer: Department of Mathematics, University of Kentucky
Location: Lexington, KY
Date: October 21, 2024
Title: "Machine Learning and Optimization in Multi-Agent Multi-Objective Frameworks"
FutureG Summer Research Camp
Organizer: DoD Center of Excellence in FutureG, Arizona State University
Location: Tempe, AZ
Date: May 7, 2024
Titles: "Overview of Reinforcement Learning", and "RL at the Edge: Fairness and Action Space Reduction"
AI-EDGE SPARKS Seminar
Organizer: The SPARKS Organizing Committee, The Ohio State University
Location: Columbus, OH
Date: March 22, 2024
Title: "Achieving Fairness in Multi-Agent MDP Using Reinforcement Learning"
The 15^th International Conference of the ERCIM WG on Computational and Methodological Statistics (CMStatistics 2022)
Organizer: King's College London
Location: London, UK
Date: 17-19 December 2022
Title: "On the Generalization Power of the Overfitted Three-Layer Neural Tangent Kernel Model"
AIRS in the AIR Academic Series
Organizer: Shenzhen Institute of Artificial Intelligence and Robotics for Society (AIRS)
Location: Shenzhen, China
Date: August 23, 2022
Title: "Distribution Level Markets under High Renewable Energy Penetration"
Long Feng Science Forum
Organizer: CUHK-Shenzhen
Location: Shenzhen, China
Date: August 19, 2022
Title: "On the Generalization Power of Overfitted Two-Layer Neural Tangent Kernel Models"
Joint Workshop between AI-EDGE and IBM
Organizer: The Ohio State University
Location: Columbus, OH
Date: June 15, 2022
Title: "On the Generalization Power of Overfitted Two-Layer Neural Tangent Kernel Models"
TDAI deep learning summer school
Organizer: The Ohio State University
Location: Columbus, OH
Date: June 3, 2022
Title: "On the Generalization Power of Overfitted Two-Layer Neural Tangent Kernel Models"

Contact

Feel free to reach out to me by email:

peizhong.ju@uky.edu

You can also find me on Google Scholar:

https://scholar.google.com/citations?user=VDzpfOYAAAAJ&hl=en

My ORCID iD:

https://orcid.org/0000-0002-4569-3539