Current Work

The primary focus of my M.Sc. research is the development of a new algorithm for modelling dynamical systems, called Compressed Predictive State Representation (or CPSR, for short). This work has two primary goals: (1) to demonstrate how dimensionality reduction techniques facilitate the efficient learning of globally optimal models via spectral methods; and (2) to develop an efficient and general model-based reinforcement learning framework that will allow agents to learn in novel environments without prior knowledge.


In addition to this main project, I am also investigating a number of related methods for modelling high-dimensional time-series data. These methods involve a wide-range of techniques, such as non-parametrics, representation learning, and spectral methods. Of primary interest to me is how to properly balance expressiveness (e.g. through Bayesian non-parametrics) with tractability (e.g. through representation learning or dimensionality reduction). I also have a keen interest in applying these methods to natural language processing/understanding problems.


Publications and Awards

Publications, Peer-Reviewed Conferences, and Works Under Review:

W. L. Hamilton, M. M. Fard, J. Pineau. "Modelling sparse dynamical systems with compressed predictive state representations". Proceedings of the Thirtieth International Conference on Machine Learning (ICML). 2013. [paper][slides]


W. L. Hamilton, M.M. Fard, J. Pineau. "Efficient Learning and Planning with Compressed Predictive States" (Extended Abstract). First Multidisciplinary Conference on Reinforcement Learning and Decision Making. 2013. [poster]


W. L. Hamilton, M.M. Fard, and J. Pineau. "Efficient Learning and Planning with Compressed Predictive States". Journal of Machine Learning Research. Submitted. [manuscript (at ArXiv)]


Awards:

McGill University Graduate Excellence Award (2013)


International Conference on Machine Learning (ICML) student travelling scholarship (2013)


Natural Sciences and Engineering Research Council (NSERC) Alexander Graham Bell Canada Graduate Scholarship (2013)


Computing Research Association Outstanding Undergraduate Researcher, Honorable Mention (2013)


McGill JW McConnell Scholarship (tuition and expenses) (2009-2013)


NSERC Undergraduate Science Research Award (2012)


McGill Undergraduate Science Research Award (2012)


McGill Computer Science Undergraduate Research Award (2012)


Oscar Wilde International Academic Gold Medal, Highly Commended (2012)


McGill Arts Research Internship Award (Cognitive Science) (2012)