I'm a computational scientist working on machine learning (a branch of Artificial Intelligence) and its application to complex biology problems. I'm especially interested in the interface between symbolic (e.g. first-order / fuzzy logic) and probabilistic approaches, and in transfer learning, which in layman's terms could be described as the capacity to automatically transfer knowledge between different tasks. It comes naturally to us, e.g. we use what we learned for walking when learning to run, but almost all machine learning algorithms build models from nothing.
I'm mostly working on:
- The interface between symbolic and probabilistic approaches (e.g. first-order logic in deep learning, Markov logic).
- The design of algorithms models capable of transferring knowledge.
- High performance scientific computing, increasingly with GPUs (CUDA/OpenCL).
- Applications to biology, most notably ecology.
I'm currently a Ph.D. candidate at the Canada Research Chair on Integrative ecology (Université de Sherbrooke), Tim Poisot's lab (Université of Montréal), and the Quebec Center for Biodiversity Science (McGill U.). My work is supported by an Alexander Graham Bell Graduate Scholarship from the NSERC and a generous Azure for Research Award from Microsoft Research. I'm a member of the Institute of Electrical and Electronics Engineers and their Computational Intelligence society. You can get my short CV here.
My work depends on many high-quality open-source software, especially the following projects: LLVM/Clang, Git, Linux, Debian/Ubuntu, Glasgow Haskell Compiler, Python, Rust, Vim, Geany, ZeroMQ, GNOME, PostgreSQL, Cassandra.