Tian Bai

Incoming Statistics Ph.D. student at Stanford University

tianbai.jpg

I recently graduated from McGill University with an Honours B.Sc. in Mathematics and Computer Science, and I will be pursuing a Ph.D. in Statistics at Stanford University starting Fall 2025. I am generally interested in statistics and statistical machine learning, and I am particularly piqued by areas such as:

  • Enhancing reliability and trustworthiness of AI/ML Systems: I am interested in the development and application of methods for accessing and improving the reliability of black-box AI/ML systems (e.g. uncertainty quantification, statistical inference with AI systems), enabling confident deployments in risk-sensitive settings.

  • Understanding foundational aspects of AI/ML Systems: I am fascinated by the inner workings of complex ML models, which are often opaque. My aim is to gain a deeper understanding of these systems or potentially enhance their transparency through structural modifications.

Currently, I am working on generalizations and applications of conformal inference under the supervision of Prof. Archer Y. Yang and also collaborating with Dr. Ying Jin.

You can download my CV here.

news

Nov 28, 2024 Vanilla conformal selection requires a predefined conformity score function, which can limit selection power by preventing score optimization without additional data splitting. Checkout OptCS, a general framework that enables statistical testing (selection) after flexible data-driven model optimization!
Nov 01, 2024 My paper comparing conformal selection with traditional methods in drug pre-screening is now available as a preprint!
Sep 28, 2024 I’m pleased to announce that our paper on using machine learning to predict patient outcomes in emergency department triage has been accepted for publication in the Canadian Journal of Emergency Medicine!
May 15, 2024 My personal website is now live.