Tiange Tim Zhou

Hi! I’m Tim. I’m a final-year undergraduate student in the combined Honours Computer Science and Statistics program at the University of British Columbia in Vancouver, BC.

My program has been incredible, but for the life of me I just can’t seem to recall what kind of snack I was holding in this picture. Please let me know if you recognize what that is.

My research interests center on application-driven machine learning (ML) methodologies, emphasizing uncertainty quantification, robustness, and computational efficiency. I am especially drawn to model applicability, human-centric design, and building trustworthy AI systems. Moreover, I am particularly interested in leveraging ML to tackle pressing environmental challenges in ecological contexts.

Piragua, piragua, new block of ice, piragua. – “In The Heights”

My current research focuses on Uncertainty Quantification in Deep Learning. Specifically, I’m developing robust and practical methods to enhance the reliability of machine learning systems used in human decision-making processes. A surprising discovery along this path led me to investigate imbalances in pairs of transfer-learned neural networks and how the scaling law for these networks differs from classical ensembling methods. I am fortunate to have the amazing Dr.Geoff Pleiss as supervisor.

Peviously, I was an Research Assistant in UBC SPIN Lab to give furry robots heartbeats and make them breathe. This project, called Happraisal, examined the impact of a person’s haptic interaction with an affective, zoomorphic robot on their ability to cognitively reappraise their emotions. As the lead developer for the robotics software, I developed real-time systems that enabled tetherless robot control and synchronized high-frequency physiological data collection from participants. I worked closely with Preeti Vyas and Bereket Guta, supervised by Professor Karon Maclean.

Outside of research, I am an enthusiastic, albeit mediocre, bird watcher, and am known to perform with UBC Acappella’s Choral Reef.