Research map / Learning theory and algorithms / Sample complexity
Learning to Guide Random Search
Ozan Sener, Vladlen Koltun
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 |
|---|---|---|---|
| learn the nonlinear search manifold online and combine manifold and full-space search | random search · MuJoCo locomotion suite | episodes to threshold: LMRS uses fewer episodes than ARS on all six tasks | Improved |
| learn the nonlinear search manifold online and combine manifold and full-space search | random search · Airfoil after 1500 XFoil simulations | Lift-Drag: 0.560 -> 1.9737 | Improved |
| add online manifold learning during search (ablation) | LMRS · MuJoCo tasks beyond Swimmer | episodes to threshold: Swimmer: 320 -> 222; Hopper: 5561 (4/5) -> 2408; HalfCheetah: 6528 -> 1128; Walker: 295493 (3/5) -> 15525; Ant: 24520 (4 | Improved |
| replace online manifold learning with offline manifold training (ablation) | LMRS · MuJoCo tasks | episodes to threshold: Swimmer: 325 -> 222; Hopper: 11616 -> 2408; HalfCheetah: 3416 -> 1128; Walker: 118640 -> 15525; Ant: 63720 (2/5) -> 9240 | Mixed |
| remove random search and optimize using surrogate gradients alone (ablation) | LMRS · MuJoCo locomotion tasks | task-solving success | Hurt |
| reduce manifold dimensionality relative to ambient dimensionality | LMRS · synthetic optimization problems | objective value | Improved |
| use LMRS rather than ARS | continuous optimization benchmarks · Pagmo suite, 46 functions | performance profile | Improved |
| use LMRS rather than REMBO | continuous control · MuJoCo locomotion suite | task-solving success | Improved |
Closest papers by meaning
- High-Dimensional Bayesian Optimization via Nested Riemannian Manifolds (NeurIPS 2020 · similarity 0.89)
- Bayesian Optimization on Function Spaces via Sparse RKHS Manifolds (NeurIPS 2026 · similarity 0.87)
- Learning-Rate-Free Stochastic Optimization over Riemannian Manifolds (ICML 2024 · similarity 0.87)
- Random Function Descent (NeurIPS 2024 · similarity 0.86)
- The Power of Random Features and the Limits of Distribution-Free Gradient Descent (ICML 2025 · similarity 0.86)
- Vanilla Bayesian Optimization Performs Great in High Dimensions (ICML 2024 · 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.
