Proposes CLARSTA, a random subspace trust-region algorithm for convex-constrained DFO with new projection-based model class, geometry measure, and concentration-of-measure subspace sampling, proving almost-sure convergence and complexity while demonstrating performance on problems up to dimension 10
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
math.OC 2verdicts
UNVERDICTED 2representative citing papers
Introduces a technique for constructing Q-fully quadratic models and applies it in a random subspace DFO algorithm with almost-sure global convergence and expected iteration bounds, plus numerical comparisons to linear models.
citing papers explorer
-
CLARSTA: A random subspace trust-region algorithm for convex-constrained derivative-free optimization
Proposes CLARSTA, a random subspace trust-region algorithm for convex-constrained DFO with new projection-based model class, geometry measure, and concentration-of-measure subspace sampling, proving almost-sure convergence and complexity while demonstrating performance on problems up to dimension 10
-
$Q$-fully Quadratic Modeling and its Application in a Random Subspace Derivative-free Method
Introduces a technique for constructing Q-fully quadratic models and applies it in a random subspace DFO algorithm with almost-sure global convergence and expected iteration bounds, plus numerical comparisons to linear models.