My research interests are in convex optimization and deep learning. Currently, I focus on the understanding of neural network through a convex perspective and on improving the speed and memory of first-order methods using sketching tools. Previously, I developed reinforcement and imitation learning algorithms for safety-critical applications.
I obtained a B.S. and M.S. at the Ecole Polytechnique, and I completed the Part III of the Mathematical Tripos at Cambridge University. I was also a visiting student researcher at UC Berkeley with Laurent El Ghaoui.
My Google Scholar and Github pages.
Email: lacotte (at) stanford (dot) edu