Welcome!
Hello, I’m Yao Zhao, a 5th-year PhD student at the University of Arizona in computer science working with Dr. Kwang-Sung Jun. I mainly have fun exploring multi-armed bandits / RL / adaptive experimentation, and recently RLHF.
I am seeking my next research position. Please contact me if you see a fit.
Selected Papers
Fixing the Loose Brake: Exponential Tail Bounds for Stopping Time in Best Arm Identification
- (co 1st author: † ) Kapilan Balagopalan†, Tuan Nguyen†, Yao Zhao†, Kwang-Sung Jun
- Under review in AISTATS
Adaptive Experimentation When You Can’t Experiment
- Yao Zhao, Kwang-Sung Jun, Tanner Fiez, Lalit Jain
- In Conference on Neural Information Processing Systems (NeurIPS), 2024, arXiv
- Also accepted and presented at Conference on Digital Experimentation (CODE@MIT), Cambridge, 2023
Revisiting Simple Regret: Fast Rates for Returning a Good Arm
- Yao Zhao, Connor Stephens, Csaba Szepesvári, Kwang-Sung Jun
- In International Conference on Machine Learning (ICML), 2023, Official
- Also accepted and presented at
- Quantifying Uncertainty: Stochastic, Adversarial, and Beyond workshop, Data-Driven Decision Processes, Simons Institute for the Theory of Computing, Berkeley, 2022
- Information Theory and Applications Workshop, San Diego, 2024
Feel free to reach out if you’d like to discuss any of these works! In a past life, I had fun with mobile network optimization and learning problems, and I published a few papers in these areas. This application journey brought me here to continue my research and exploration on a deeper and more theoretical level.
Industry Experiences
Applied Scientist Intern, Amazon Summer 2023, summer 2024
- Mentors: Dr. Lalit Jain, Dr. Tanner Fiez, Dr. Bibek Adhikari
- Project blog: Adaptive experimentation when you can’t experiment
- Worked with cross-functional teams of ML scientists, economists.
- Developed a causal inference ML solution for optimizing Amazon’s membership tier under confounding and compliance issues to support key products decision making.
- Developed a LLM augmented experimentation solution.
Teaching Experiences
Principles of Data Science (Undergrad-level class with 100+ students) Fall, 2024
Algorithms (Undergrad-level class with 60 students) Spring, 2023
Machine Learning (Grad-level class with 100+ students, Excellent Teaching Assistant Award) Spring, 2019
Probability and Statistics (Undergrad-level class with 100+ students) Spring, 2018
Professional Services
I’ve served as a reviewer for major ML conferences, including ICML, NeurIPS, AISTATS, ICLR etc.
Award
NeurIPS Scholar Award 2024
Travel Award for ICML 2023
Graduate College Fellowship 2021
Excellent Teaching Assistant Award 2019