Han Shao

Email: [first name] at ttic.edu.

Hi! I am a CMSA postdoc at Harvard, working with Cynthia Dwork and Ariel Procaccia. Starting in fall 2025, I will be an Assistant Professor in the Department of Computer Science at the University of Maryland, College Park (UMD). Before that, I did my PhD at TTIC,  where I was extremely fortunate to be advised by Avrim Blum

I am hiring PhD students at UMD CS. If you're interested, apply to the program and mention me as the faculty member of interest. If your interests and experience strongly align with mine, please feel free to email me your CV, research interests, and how your background connects with my research.

My primary research centers on machine learning theory, with a specific focus on modeling human strategic and adversarial behaviors within the learning process. I aim to understand how these behaviors affect machine learning systems and develop methods to enhance accuracy/robustness. Additionally, I am interested in gaining a theoretical understanding of empirical observations concerning adversarial robustness. 

Publications

(α-β denotes alphabetical ordering, * denotes equal contribution)

Preprints

On the Effect of Defections in Federated Learning and How to Prevent Them

(α-β) Minbiao Han, Kumar Kshitij Patel, Han Shao, and Lingxiao Wang

arxiv, 2023.

Conference Papers

Efficient Prior-Free Mechanisms for No-Regret Agents

(α-β) Natalie Collina, Aaron Roth, and Han Shao

EC, 2024.

Learnability Gaps of Strategic Classification

(α-β) Lee Cohen, Yishay Mansour, Shay Moran, and Han Shao

COLT, 2024.

Incentivized Collaboration in Active Learning

(α-β) Lee Cohen and Han Shao

FORC, 2024.

Strategic Classification under Unknown Personalized Manipulation

Han Shao, Avrim Blum and Omar Montasser

NeurIPS, 2023.

Eliciting User Preferences for Personalized Multi-Objective Decision Making through Comparative Feedback

Han Shao, Lee Cohen, Avrim Blum, Yishay Mansour, Aadirupa Saha and Matthew R. Walter

NeurIPS, 2023.

A Theory of PAC Learnability under Transformation Invariances

Han Shao, Omar Montasser and Avrim Blum

NeurIPS, 2022. (Oral)

Robust Learning under Clean-Label Attack

(α-β) Avrim Blum, Steve Hanneke, Jian Qian and Han Shao

COLT,  2021. 

One for One, or All for All: Equilibria and Optimality of Collaboration in Federated Learning

(α-β) Avrim Blum, Nika Haghtalab, Richard Lanas Phillips and Han Shao

ICML, 2021

Online learning with preliminary and secondary losses

(α-β) Avrim Blum and Han Shao

NeurIPS, 2020.

Structure adaptive algorithms for stochastic bandits

Rémy Degenne, Han Shao and Wouter M. Koolen

ICML, 2020.

Almost optimal algorithms for linear stochastic bandits with heavy-tailed payoffs

Han Shao*, Xiaotian Yu*, Irwin King, and Michael R. Lyu

NeurIPS, 2018. (Spotlight)

Pure exploration of multi-armed bandits with heavy-tailed payoffs

Xiaotian Yu, Han Shao, Michael R. Lyu, and Irwin King

UAI, 2018.

Awards

EECS Rising Star, Georgia Tech, 2023

Rising Star in Machine Learning, University of Maryland, 2023

Services

Co-organizer of the TTIC Student Workshop 2024.

Student member of TTIC DEI committee (2023-2024).