Research map / Learning theory and algorithms / Matrix and tensor methods
Inexact trust-region algorithms on Riemannian manifolds
Hiroyuki Kasai, Bamdev Mishra
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| sub-sample the Hessian only (Sub-H-RTR) instead of full-Hessian RTR | Riemannian trust-region · PCA (Cases P1-P4) | optimality gap vs run time | Improved |
| also sub-sample the gradient (Sub-HG-RTR) on top of sub-sampled Hessian (ablation) | Sub-H-RTR · PCA synthetic (Cases P1-P2) | optimality gap vs run time | Hurt |
| vary Hessian sample size |S_H| among n/10, n/100, n/1000 (ablation) | Sub-H-RTR · PCA Case P1 | optimality gap vs run time | No clear effect |
| use fixed sample size instead of linear/adaptive growth (ablation) | Sub-H-RTR and Sub-HG-RTR · MC Cases M4-M5 | test MSE vs run time | Improved |
| fixed sample size for Sub-HG-RTR versus original RTR (ablation) | Sub-HG-RTR · PCA Case P6 | optimality gap vs run time | Hurt |
| sub-sampled RTR versus RSVRG | Sub-H-RTR / Sub-HG-RTR · ICA (Cases I1-I3) | cost vs oracle calls/run time | Mixed |
| sub-sampled RTR versus first-order and RTR baselines on ill-conditioned matrix completion | Sub-H-RTR / Sub-HG-RTR · MC synthetic CN=20 (Case M2) | test MSE | Improved |
Closest papers by meaning
- Efficient Optimization with Orthogonality Constraint: a Randomized Riemannian Submanifold Method (ICML 2025 · similarity 0.87)
- Convergence and Complexity Guarantee for Inexact First-order Riemannian Optimization Algorithms (ICML 2024 · similarity 0.86)
- Decentralized Projected Riemannian Stochastic Recursive Momentum Method for Nonconvex Optimization (AAAI 2025 · similarity 0.86)
- Implicit Riemannian Optimism with Applications to Min-Max Problems (ICML 2025 · similarity 0.86)
- Nonconvex Federated Learning on Compact Smooth Submanifolds With Heterogeneous Data (NeurIPS 2024 · similarity 0.85)
- SERENA: A Unified Stochastic Recursive Variance Reduced Gradient Framework for Riemannian Non-Convex Optimization (ICML 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
