Nicole Dumont

PhD Student

I joined the CRNG lab in September 2019 as a PhD student in Computer Science and was co-supervised by Jeff Orchard from the Neurocognitive Computing Lab. The focus of my thesis was modelling spatial cognition in the brain with neurosymbolic methods. I hypothesized that particular neurosymbolic methods -- namely, Holographic Reduced Representations (Plate, 1995) extended to include dynamics -- were a good model of spatial encoding in the medial entorhinal cortex. These representations can be used to model grids cells, path integration, cognitive mapping, and SLAM. I additionally studied biologically-plausible reinforcement learning and successor representations.

I obtained a M.Math in the computational mathematics program at the University of Waterloo, where I researched continuous optimization and was advised by Thomas Coleman. I obtained my B.Sc. in mathematics and physics from McMaster University.

Publications

Journal Articles

Conference and Workshop Papers