Kai
Wang

General Information

Email:
kwang692@gatech.edu
Phone:
6265888197
Location - Building:
Coda
Location - Room:
S1309
Roles:
Professor (any rank)
Primary Unit:
School of Computational Science and Engineering
Website URL
https://guaguakai.com/

Details

Degrees with subject and Postdoc Experience:
Degree Type
Ph.D.
Subject
Computer Science
Year
2023
Institution
Harvard University
Location
Cambridge, MA
Degree Type
Postdoctoral scholar
Subject
Computer Science
Year
2023 - 2024
Institution
Massachusetts Institute of Technology
Location
Cambridge, MA
Degree Type
B.S.
Subject
Electrical Engineering, Mathematics
Year
2017
Institution
National Taiwan University
Location
Taipei, Taiwan
Statement of Research Interests:

Dr. Kai Wang’s research develops the computational foundations of AI for healthcare, creating scalable machine learning and optimization methods for high-stakes decision-making. His work pioneers "Generative Decision Intelligence" to transform generative models into actionable decision agents and advances Decision-Focused Learning to align predictive models directly with optimization objectives. To enable these systems at scale, he has introduced novel fully first-order algorithms for bilevel optimization that resolve critical computational bottlenecks in high-dimensional decision spaces. Integrating these methods with sequential decision-making frameworks like Restless Multi-Armed Bandits and Online/Offline Reinforcement Learning, he leads interdisciplinary teams to deploy AI solutions for maternal health, diabetes management, and pediatric critical care in partnership with institutions such as Children’s Healthcare of Atlanta and Massachusetts General Hospital. Kai’s work has been recognized with the Schmidt Science AI2050 Early Career Fellowship, the Siebel Scholarship, and the Best Paper Runner-Up Award at AAAI.

Statement of Teaching Interests:

Dr. Kai Wang’s teaching interests lie at the intersection of rigorous computational theory and its application to high-impact societal domains, particularly healthcare. He focuses on instructing students in the mathematical foundations of machine learning—including optimization, deep learning, and generative models—while developing curricula that bridge the gap between abstract theory and practical deployment, as exemplified by courses like Computational Data Analysis (CSE 6740) and AI for Social Impact (CSE 8803). Additionally, he is interested in cultivating an inclusive, research-driven environment that encourages students to connect these computational tools with their own research inquiries into critical real-world challenges.

Selection of recent research, scholarly, and creative activities:

Dr. Kai Wang’s research advances the computational foundations of AI for social impact, with a primary focus on healthcare and public health. In 2026, he was selected for the AAAI New Faculty Highlight Program, where he presented his vision for decision-focused AI. He also served as a tutorial organizer at AAAI 2026, delivering a session on "Generative AI in Healthcare: Causality, Decision, and Real-world Case Study."

Preprints

A Fully First-Order Layer for Differentiable Optimization
Zihao Zhao, Kai-Chia Mo, Shing-Hei Ho, Brandon Amos, Kai Wang (Github link, in submission)

Finding a Multiple Follower Stackelberg Equilibrium: A Fully First-Order Method
April Niu*, Kai Wang*, Juba Ziani* (in submission)

One-Step Flow Policy Mirror Descent
Tianyi Chen, Haitong Ma, Na Li, Kai Wang*, Bo Dai* (in submission)

Non-Stationary Restless Multi-Armed Bandits with Provable Guarantee
Yu Heng Hung, Kai Wang, Ping-Chun Hsieh (in submission)

Neural Index Policies for Restless Multi-Action Bandits with Heterogeneous Budgets
Himadri S Pandey, Kai Wang, Gian-Gabriel P. Garcia (in submission)

Networked Restless Multi-Arm Bandits with Reinforcement Learning
Hanmo Zhang, Zenghui Sun, Kai Wang (PRL workshop AAAI 2025, in submission)

 

Publications in 2025 - 2026

Diffusion-DFL: Decision-focused Diffusion Models for Stochastic Optimization
Zihao Zhao, Christopher Yeh, Lingkai Kong, Kai Wang (ICLR 2026)

Efficient Online Reinforcement Learning for Diffusion Policy
Haitong Ma, Tianyi Chen, Kai Wang, Li Na*, Bo Dai* (ICML 2025)

Opportunistic Screening of Type 2 Diabetes with Deep Metric Learning using Electronic Health Records
Qixuan Jin, Haoran Zhang, Lukasz Szczerbinski, Jiacheng Zhu, Walter Gerych, Xuhai Xu, Kai Wang, Sarah Hsu, Ravi Mandla, Aaron Deutsch, Alisa Manning, Josep Mercader, Thomas Hartvigsen, Miriam Udler, Marzyeh Ghassemi (Scientific Reports 2025)

Primal-Dual Spectral Representation for Off-policy Evaluation
Yang Hu, Tianyi Chen, Na Li, Kai Wang, Bo Dai (AISTATS 2025)

What is the Right Notion of Distance between Predict-then-Optimize Tasks?
Paula Rodriguez-Diaz, Lingkai Kong, Kai Wang, David Alvarez-Melis, Milind Tambe (UAI 2025)

What’s in a Query: Polarity-aware Distribution-based Fair Ranking
Aparna Balagopalan*, Kai Wang*, Olawale Elijah Salaudeen, Asia Biega, Marzyeh Ghassemi (WWW 2025)