About me.
About Me
Hi! I'm Xavier, a published mathematician with a background in classical music, now focused on AI research. I’m excited to be working on time series modelling and uncertainty quantification, with an particular focus on applications to healthcare. If you’re interested in a quick summary of my research, see directly below, and more information about my background can be found at the bottom of the page.
-
Data in the real world is distributed through time. Time series analysis isn’t a new idea, dating back to the era of Galton (1880s), but nonetheless, modelling time series remains a significant challenge in deep learning. Primarily I focus on irregular time series, which may be irregularly-sampled or variable-length, in which the traditional deep learning paradigms fail. One of the problems I focus on now is seizure onset zone (SOZ) localization, which is the task of identifying which region of the brain is contributing to seizures in epilepsy patients. Traditionally, neurosurgeons and epileptologists visually inspect intracranial EEG signals to determine the SOZ. However, this challenge isn't so easy to apply any off-the-shelf deep learning algorithm. The signals can be extremely long (due to high sampling rate), and variable in their length (for different recordings or patients). We have some exciting work that will be releasing shortly on this in the next month. Other interests include multimodal time series, where each channel may represent a completely different (but related) phenomena.
-
-
My background
Currently, I am a MSc in Computer Science student at the Lassonde School of Engineering, York University within the NWGNR Lab. My thesis focuses on self-supervised learning (SSL), time series forecasting and classification, and multimodal learning. I also work in parallel with the NSBPSL within the Krembil Research Institute at Toronto Western Hospital with a focus on developing novel deep learning methods for interpretation of neurological diseases and their clinical applications. Last summer, I was a research intern at T-CAIREM, advancing SSL techniques with GNN encoders to boost seizure detection downstream, for which I am continuing my work on seizure onset zone (SOZ) localization, which involves locating the region of the brain believed to be contributing to seizures. During my undergrad I had the privilege of holding the NSERC USRA position twice at York University where I worked on algebraic combinatorics and operator algebras. I also have a fascination with quantum computing, which led me to participate in the esteemed USEQIP program at the Institute for Quantum Computing, University of Waterloo. Before all of this, my research journey began at the Population Neuroscience & Developmental Neuroimaging Lab within Holland Bloorview Kids Rehabilitation Hospital.
-
MSc in Computer Science — York University (2023—2025)
Specialized Honours BSc in Mathematics — York University (2019—2023)
Music Industry Arts — Fanshawe College (2017—2019)
Schulich School of Music — McGill University (2013—2015)