C
Chris Finlay
Researcher at McGill University
Publications - 31
Citations - 514
Chris Finlay is an academic researcher from McGill University. The author has contributed to research in topics: Robustness (computer science) & Solver. The author has an hindex of 10, co-authored 31 publications receiving 349 citations. Previous affiliations of Chris Finlay include University of Alberta.
Papers
More filters
Posted Content
How to train your neural ODE: the world of Jacobian and kinetic regularization
TL;DR: This paper introduces a theoretically-grounded combination of both optimal transport and stability regularizations which encourage neural ODEs to prefer simpler dynamics out of all the dynamics that solve a problem well, leading to faster convergence and to fewer discretizations of the solver.
Proceedings Article
How to Train Your Neural ODE: the World of Jacobian and Kinetic Regularization
TL;DR: In this article, a theoretically-grounded combination of both optimal transport and stability regularization is proposed to encourage neural ODEs to prefer simpler dynamics out of all the dynamics that solve a problem well.
Posted Content
Lipschitz regularized Deep Neural Networks generalize and are adversarially robust
TL;DR: This work studies input gradient regularization of deep neural networks, and demonstrates that such regularization leads to generalization proofs and improved adversarial robustness.
Journal ArticleDOI
Are more complicated tumour control probability models better
TL;DR: A systematic study where six TCP models are compared, including the Poissonian TCP, the Zaider-Minerbo TCP, a Monte Carlo TCP and their corresponding cell cycle (two-compartment) models, and it is found that in realistic treatment scenarios, all one- and two-compartments TCP models give basically the same results.
Posted Content
How to train your neural ODE
TL;DR: In this paper, a theoretically-grounded combination of both optimal transport and stability regularization is proposed to encourage neural ODEs to prefer simpler dynamics out of all the dynamics that solve a problem well.