A KL-divergence trust-region formulation for sampling-based MPC replaces heuristic hyperparameter adaptation with Lagrangian-optimal updates and improves convergence when combined with deterministic LCD sampling.
A randomized Halton algorithm in R
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
abstract
Randomized quasi-Monte Carlo (RQMC) sampling can bring orders of magnitude reduction in variance compared to plain Monte Carlo (MC) sampling. The extent of the efficiency gain varies from problem to problem and can be hard to predict. This article presents an R function rhalton that produces scrambled versions of Halton sequences. On some problems it brings efficiency gains of several thousand fold. On other problems, the efficiency gain is minor. The code is designed to make it easy to determine whether a given integrand will benefit from RQMC sampling. An RQMC sample of n points in $[0,1]^d$ can be extended later to a larger n and/or d.
citation-role summary
dataset 1
citation-polarity summary
verdicts
UNVERDICTED 3roles
dataset 1polarities
use dataset 1representative citing papers
Labrador is a domain-optimized neural posterior estimation tool achieving 1% median importance-sampling efficiency and first extensive coverage of long-duration low-mass gravitational wave signals through equivariance and a stable procedure for differing priors.
Mean dimension of ridge functions is bounded for Lipschitz cases as d grows to infinity, scales as O(sqrt(d)) for discontinuous non-sparse cases, and preintegration reduces it to O(1) under a non-vanishing coefficient condition.
citing papers explorer
-
Sampling-based Model Predictive Control Using Trust Regions
A KL-divergence trust-region formulation for sampling-based MPC replaces heuristic hyperparameter adaptation with Lagrangian-optimal updates and improves convergence when combined with deterministic LCD sampling.
-
labrador: A domain-optimized machine-learning tool for gravitational wave inference
Labrador is a domain-optimized neural posterior estimation tool achieving 1% median importance-sampling efficiency and first extensive coverage of long-duration low-mass gravitational wave signals through equivariance and a stable procedure for differing priors.
-
Mean Dimension of Ridge Functions
Mean dimension of ridge functions is bounded for Lipschitz cases as d grows to infinity, scales as O(sqrt(d)) for discontinuous non-sparse cases, and preintegration reduces it to O(1) under a non-vanishing coefficient condition.