Paper ID: 2111.04596
Inertial Newton Algorithms Avoiding Strict Saddle Points
Camille Castera
We study the asymptotic behavior of second-order algorithms mixing Newton's method and inertial gradient descent in non-convex landscapes. We show that, despite the Newtonian behavior of these methods, they almost always escape strict saddle points. We also evidence the role played by the hyper-parameters of these methods in their qualitative behavior near critical points. The theoretical results are supported by numerical illustrations.
Submitted: Nov 8, 2021