I'm interested in
Mathematical Optimization and Machine Learning
Zeroth order optimization (Black-box optimization, Derivative-free optimization)
Newton and Quasi-Newton type methods
Structured DFO (Derivative-Free Optimization)
Mean-Field Games
Variance Reduction
Black Box Adversarial attack
Topic Modeling
Papers
Bumsu Kim, HanQin Cai, Daniel McKenzie, and Wotao Yin. Curvature-Aware Derivative-Free Optimization, submitted. (ArXiv) (GitHub)
Bumsu Kim and Wotao Yin. Bridging the Gap Between Local and Global Derivative-Free Optimization, in preparation.
Talks
ICCOPT 2022, July 2021, Curvature-Aware Random Search
INFORMS 2021, October 2021. Curvature-Aware Derivative-Free Optimization
Zeroth Order Online Meeting, July 2020. Zeroth order Newton-type algorithms with low rank Hessian updates
Interesting Stuffs
For those who are interested in acceleration methods, I would recommend watching this talk by Ernest Ryu: "Non-Nesterov Acceleration Methods for Minimization and Minimax Optimization" from One World Optimization Seminar.