Timezone: »

 
Poster
Inexact trust-region algorithms on Riemannian manifolds
Hiroyuki Kasai · Bamdev Mishra

Wed Dec 05 07:45 AM -- 09:45 AM (PST) @ Room 210 #15
in Wed Poster Session A »

We consider an inexact variant of the popular Riemannian trust-region algorithm for structured big-data minimization problems. The proposed algorithm approximates the gradient and the Hessian in addition to the solution of a trust-region sub-problem. Addressing large-scale finite-sum problems, we specifically propose sub-sampled algorithms with a fixed bound on sub-sampled Hessian and gradient sizes, where the gradient and Hessian are computed by a random sampling technique. Numerical evaluations demonstrate that the proposed algorithms outperform state-of-the-art Riemannian deterministic and stochastic gradient algorithms across different applications.

Keywords: Nonlinear Dimensionality Reduction and Manifold Learning · Non-Convex Optimization
[ Paper]  [ Poster] 

Author Information

Hiroyuki Kasai (The University of Electro-Communications)
Bamdev Mishra (Microsoft)

More from the Same Authors