Research

On This Page

I publish under the name Xu Kuang, cited as X. Kuang. Prior to June 2023, I published under the name Kuang Xu, cited as K. Xu.

Working Papers

Non-Stationary Contextual Bandit Learning via Neural Predictive Ensemble Sampling
Co-authored with: Yueyang Liu, Benjamin van Roy, Zheqing (Bill) Zhu

A Definition of Non-Stationary Bandits
Co-authored with: Yueyang Liu, Benjamin van Roy

Experimenting under Stochastic Congestion
Co-authored with: Shuangning Li, Ramesh Johari, Stefan Wager

Detecting Service Slowdown using Observational Data
Co-authored with: Gal Mendelson

Treatment Effects in Market Equilibrium
Co-authored with: Evan Munro, Stefan Wager

CARE: Resource Allocation Using Sparse Communication
Co-authored with: Gal Mendelson

Nonstationary Bandit Learning via Predictive Sampling
Co-authored with: Yueyang Liu, Benjamin van Roy
Preliminary version, AISTATS 2023. 

Publications

Journal Publications

Weak Signal Asymptotics for Sequentially Randomized Experiments
Co-authored with: Stefan Wager
Forthcoming in Management Science.

Learner-Private Online Convex Optimization
Co-authored with: Jiaming Xu, Dana (Xiaoqian) Yang
IEEE Transactions on Information Theory, 2022, 69, 528-547.

Anonymous Stochastic Routing
Co-authored with: Mine Su Erturk
Stochastic Systems, 2022, 12.1, 92-126.
arXiv version
Finalist, INFORMS George E. Nicholson Prize, 2019. 

Learning and Information in Stochastic Networks and Queues
Co-authored with: Neil Walton
INFORMS TutORials in Operations Research, 2021, 161-198.
arXiv version

Private Sequential Learning
Co-authored with: John N. Tsitsiklis, Zhi Xu
Operations Research, 2021, 69, 5:1575-1590. 

Temporal Concatenation for Markov Decision Processes
Co-authored with: Ruiyang Song
Probability in the Engineering and Informational Sciences, 2021, 1-28.
arXiv version

Experimenting in Equilibrium
Co-authored with: Stefan Wager
Management Science, 2021, 67, 11: 6694-6715.
arXiv version

Information and Memory in Dynamic Resource Allocation  
Co-authored with: Yuan Zhong
Operations Research, 2020, 68.6: 1625-1931. 

Reinforcement with Fading Memories
Co-authored with: Se Young Yun
Mathematics of Operations Research, 2020, 45.4: 1193-1620. 

No Detectable Surge in SARS-CoV-2 Transmission due to the April 7, 2020 Wisconsin Election
Co-authored with: Kathy Leung, Joseph T. Wu, Lawrence M. Wein
American Journal of Public Health, 2020, 1123–1172

Delay-Predictability Tradeoffs in Reaching a Secret Goal 
Co-authored with: John N. Tsitsiklis
Operations Research, 2018, 66.2: 587-596.
SSRN version

On the Capacity of Information Processing Systems
Co-authored with: Laurent Massoulié
Operations Research, 66(2): 568-586, 2018.

Flexible Queueing Architectures
Co-authored with: John N. Tsitsiklis
Operations Research, 2017, 65.5: 1398–1413.
arXiv version

Using Future Information to Reduce Waiting Times in the Emergency Department via Diversion
Co-authored with: Carri W. Chan
Manufacturing & Service Operations Management (MSOM), 2016, 18.3: 314 – 331.

Necessity of Future Information in Admission Control 
Operations Research, 63(5): 1213-1226, 2015.
Manuscript

Queueing with Future Information
Co-authored with: Joel Spencer, Madhu Sudan
Annals of Applied Probability, 2014, 24.5: 2091-2142.
arXiv version

The Optimal Admission Threshold in Observable Queues with State Dependent Pricing
Co-authored with: Christian Borgs, Jennifer T. Chayes, Sherwin Doroudi, Mor Harchol-Balter
Probability in the Engineering and Informational Sciences, 28(1): 101-119, 2014. 
Manuscript

On the Power of (even a little) Resource Pooling
Co-authored with: John N. Tsitsiklis
Stochastic Systems, 2012, 2: 1-66.  
First Place, INFORMS George E. Nicholson Prize, 2021. 

Conference Publications

Nonstationary Bandit Learning via Predictive Sampling
Co-authored with: Yueyang Liu, Benjamin van Roy
Proceedings of International Conference on Artificial Intelligence and Statistics (AISTATS), 2023.

Learner-Private Online Convex Optimization
Co-authored with: Jiaming Xu, Dana (Xiaoqian) Yang
Proceedings of International Conference on Machine Learning (ICML) 2021.

Optimal Query Complexity for Private Sequential Learning against Eavesdropping
Co-authored with: Jiaming Xu, Dana (Xiaoqian) Yang
Proceedings of International Conference on Artificial Intelligence and Statistics (AISTATS) 2021.
arXiv version

Information, Memory and Capacity in Dynamic Resource Allocation
Co-authored with: Yuan Zhong
Proceedings of ACM SIGMETRICS,Phoenix, Arizona, June 2019.
Extended technical report

Query Complexity of Bayesian Private Learning
Proceedings of the Conference on Neural Information Processing Systems (NeurIPS), Montreal, December 2018.
arXiv version

Private Sequential Learning
Co-authored with: John N. Tsitsiklis, Zhi Xu
Proceedings of the Conference on Learning Theory (COLT), Stockholm, July 2018.

Reinforcement with Fading Memories
Co-authored with: Se Young Yun
Proceedings of ACM SIGMETRICS, Irvine, California, June 2018.

On the Capacity of Information Processing Systems
L. Massoulié and K. Xu, Proceedings of the Conference on Learning Theory (COLT) , New York, June 2016.

Queueing System Topologies with Limited Flexibility
Co-authored with: John N. Tsitsklis
Proceedings of ACM SIGMETRICS , Pittsburg, June 2013.
Best Paper Award and Kenneth C. Sevcik Outstanding Student Paper Award
Extended technical report

On the Power of (even a little) Centralization in Distributed Processing
Co-authored with: John N. Tsitsklis
Proceedings of ACM SIGMETRICS , San Jose, June 2011.

Self-Synchronizing Properties of CSMA Wireless Multi-hop Networks 
Co-authored with: Olivier Dousse, Patrick Thiran, 
Proceedings of ACM SIGMETRICS , New York, June 2010.

PAPR Reduction for Beamforming OFDM via Constellation-Beam Modification
Co-authored with: Douglas L. Jones
Proceedings of the 43rd Asilomar Conference on Signals, Systems and Computers, Pacific Grove, November 2009.

Promoting undergraduate research in ECE (PURE): Connecting undergraduates with graduate research mentors
Co-authored with: Elizabeth Van Ruitenbeek, 
Proceedings of American Society for Engineering Education Conference & Exposition, June 2009.  

Conference Publications

Theses

On the Power of (even a little) Flexibility in Dynamic Resource Allocation
Ph.D. Thesis, Massachusetts Institute of Technology, June 2014. 
Recipient of 2014 Dimitris N. Chorafas Foundation Award

On the Power of Centralization in Distributed Processing
S.M. Thesis, Massachusetts Institute of Technology, June 2011. 
First Place, Ernst A. Guillemin Thesis Award for Best S.M. Thesis in Electrical Engineering

Modeling and Analysis of Peer-to-Peer (P2P) Live Video Streaming
Undergraduate Thesis, University of Illinois at Urbana-Champaign, May 2009. 

Other Manuscripts

Gaussian Imagination in Bandit Learning
Co-authored with: Yueyang Liu, Aditya Devraj, Benjamin van Roy

A bit better? Quantifying information for bandit learning  [arXiv]
Co-authored with: Aditya Devraj, Benjamin van Roy

Private Genetic Genealogy Search  [SSRN] [Podcast] [GSB News] M. S. Erturk and K. Xu, manuscript (2022).

Beamforming MISO-OFDM PAPR reduction: A Space-user perspective  [paper]
K. Xu, T. Thomas, and D. L. Jones, Technical Report, 2009.

Media on My research

Talks

Information-centric thinking in stochastic modeling and decision-making
ACM SIGMETRICS Rising Star Research Award Lecture, 2020

Privacy sequential learning
COLT 2018

On the capacity of information processing system
COLT 2016

On the power of (even a little) resource pooling
Cornell ORIE Colloquium, 2013