Nakul Upadhya

Research

Hi! I’m Nakul Upadhya and I’m a Doctoral Candidate at the Optimization and Machine Learning (OptiMaL) Lab at the University of Toronto under Professor Eldan Cohen. My research focuses on developing inherently interpretable machine learning models for tabular and time-series data, with applications in manufacturing and healthcare. My current line of work focuses on the concept of augmenting interpretable ML algorithms with “shape functions” - feature specific non-linear transformations - that allow for increased model expressiveness while retaining interpretability.

Selected Publications

Away from the desk

Apart from research, I’m very interested in speciality coffee and trying new coffee beans and cafes. Check out the coffee I’m drinking! Currently I’m developing a machine learning model for predicting espresso grind sizes for various coffees. If you are interested in contributing data for this project, send me an email! Additionally, I’m getting back into actual reading apart from papers, so check out my recently read books.