Tianyu Wang

I'm Tianyu Wang. I'm a PhD candidate at Duke CS advised by Prof. Cynthia Rudin. My research focuses on bandits and reinforcement learning. In the long run, I'm particularly interested in

  1. data-driven decision making with singuralities in the domain (P1).

This problem is motivated by important real-world applications, namely

  1. identify/quantify/handle singularities and instabilities in AI systems.

To this end, my research centers around principled methods in machine learning, uses some concepts in optimization, and touches upon applied classical analysis.

My email address is tianyu AT cs DOT duke DOT edu.

Some recent publications/preprints

Bandits for BMO Functions, ICML, 2020 (bib|pdf), note: related to (P1)

Tianyu Wang, Cynthia Rudin