معرفی
Subha Maity is an Assistant Professor in the Department of Statistics & Actuarial Science at the University of Waterloo. Their research focuses on statistical foundations of data science, including large language models, distribution shift, performative prediction, fairness in ML, and missing data mechanisms. They hold a PhD from the University of Michigan (advised by Prof. Yuekai Sun and Prof. Moulinath Banerjee) and completed undergraduate/masters studies at the Indian Statistical Institute, Kolkata.
- Education:
- PhD in Statistics, University of Michigan
- Master's & Undergraduate, Indian Statistical Institute, Kolkata (Mathematics/Statistics)
Recent work includes CARROT router for LLM optimization (arXiv 2024), AISTATS 2025 papers on strategic prediction and performative prediction, and NeurIPS 2024 contributions on weak supervision evaluation. Teaching includes computational statistics (Winter 2025) and undergraduate statistics courses since 2018.
Labs/Teams: Active in developing benchmarks like SPROUT dataset and releasing models via HuggingFace community contributions.
