I am a postdoctoral fellow at Simons Institute for the Theory of Computing, hosted by Bin Yu and Peter Bartlett. Before joining UC Berkeley, I spent two years working in Apple AIML Evaluation team after completing my PhD at UIUC. I was advised by Sanmi Koyejo and worked closely with Mladen Kolar.
My PhD thesis focuses on estimations and predictions in dynamic systems where data are collected in non-i.i.d. fashion. My PhD research was supported by NSF Graduate Research Fellowship.
More recently, I am interested in AI evaluation and the dynamics of the machine learning development cycle.
High-dimensional Markov-switching Ordinary Differential Processes
[ arXiv ]
Katherine Tsai, Mladen Kolar, Sanmi Koyejo
2024+
Latent Multimodal Functional Graphical Model Estimation
[ manuscript ] [ code ] [ arXiv ] [ poster ]
Katherine Tsai, Boxin Zhao, Sanmi Koyejo, Mladen Kolar
Journal of the American Statistical Association (JASA), 2024
A Nonconvex Framework for Structured Dynamic Covariance Recovery
[ manuscript ] [ code ] [ arXiv ] [ poster ]
Katherine Tsai, Mladen Kolar, Sanmi Koyejo
Journal of Machine Learning Research (JMLR), 2022
Proxy Methods for Domain Adaptation
[manuscript] [code] [ arXiv ] [ poster ]
Katherine Tsai, Stephen R. Pfohl, Olawale Salaudeen, Nicole Chiou, Matt J. Kusner, Alexander D'Amour, Sanmi Koyejo, Arthur Gretton
International Conference on Artificial Intelligence and Statistics (AISTATS), 2024
Adapting to Latent Subgroup Shifts via Concepts and Proxies
[ manuscript ] [ code ] [ arXiv ]
Ibrahim Alabdulmohsin, Nicole Chiou, Alexander D'Amour, Arthur Gretton, Sanmi Koyejo, Matt J. Kusner, Stephen R. Pfohl, Olawale Salaudeen, Jessica Schrouff, Katherine Tsai (in alphabetical order)
International Conference on Artificial Intelligence and Statistics (AISTATS), 2023
Let's chat