A Derivative-Free Trust-Region Algorithm for the Optimization of Functions Smoothed via Gaussian Convolution Using Adaptive Multiple Importance Sampling

DOI10.1137/15M1031679zbMath1390.90410OpenAlexW2605076685WikidataQ129869394 ScholiaQ129869394MaRDI QIDQ4641668

Irina S. Dolinskaya, Jeremy Staum, Alvaro Maggiar, Andreas Wächter

Publication date: 18 May 2018

Published in: SIAM Journal on Optimization (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1137/15m1031679

zbMATH Keywords

Monte Carlo samplingderivative-free optimizationtrust-region methodGaussian smoothingadaptive multiple importance samplingdeterministic computational noise

Mathematics Subject Classification ID

Nonlinear programming (90C30) Derivative-free methods and methods using generalized derivatives (90C56) Stochastic programming (90C15)

Related Items (14)

Gradient and diagonal Hessian approximations using quadratic interpolation models and aligned regular bases ⋮ A theoretical and empirical comparison of gradient approximations in derivative-free optimization ⋮ Finite Difference Gradient Approximation: To Randomize or Not? ⋮ Robust design optimization for enhancing delamination resistance of composites ⋮ Two-stage nested simulation of tail risk measurement: a likelihood ratio approach ⋮ Limiting behaviour of the generalized simplex gradient as the number of points tends to infinity on a fixed shape in \(\mathrm{IR}^n\) ⋮ Derivative-Free Optimization of Noisy Functions via Quasi-Newton Methods ⋮ Bilevel parameter learning for nonlocal image denoising models ⋮ A stochastic subspace approach to gradient-free optimization in high dimensions ⋮ Asymptotics of the optimal value of SAA with AMIS on minimax stochastic programs ⋮ Adaptive regularization for nonconvex optimization using inexact function values and randomly perturbed derivatives ⋮ Derivative-free optimization methods ⋮ Global Convergence Rate Analysis of a Generic Line Search Algorithm with Noise ⋮ Linesearch Newton-CG methods for convex optimization with noise

Uses Software

• DFO
• WEDGE
• IMFIL
• KELLEY
• minpack
• UOBYQA
• ASTRO-DF

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Stochastic derivative-free optimization using a trust region framework
• Computation of sparse low degree interpolating polynomials and their application to derivative-free optimization
• Stochastic optimization using a trust-region method and random models
• UOBYQA: unconstrained optimization by quadratic approximation
• Probability essentials.
• The effect of deterministic noise in subgradient methods
• Geometry of interpolation sets in derivative free optimization
• Convergence of Trust-Region Methods Based on Probabilistic Models
• Estimating Derivatives of Noisy Simulations
• Self-Correcting Geometry in Model-Based Algorithms for Derivative-Free Unconstrained Optimization
• Estimating Computational Noise
• Adaptive Multiple Importance Sampling
• Implicit Filtering
• Interaction of finite-amplitude waves with vertically sheared current fields
• On the geometry phase in model-based algorithms for derivative-free optimization
• Introduction to Derivative-Free Optimization
• `` Direct Search Solution of Numerical and Statistical Problems
• Testing Unconstrained Optimization Software
• Trust Region Methods
• Safe and Effective Importance Sampling
• ASTRO-DF: A Class of Adaptive Sampling Trust-Region Algorithms for Derivative-Free Stochastic Optimization
• Advances and Trends in Optimization with Engineering Applications
• Adaptative Monte Carlo Method, A Variance Reduction Technique
• Benchmarking Derivative-Free Optimization Algorithms
• Global Convergence of General Derivative-Free Trust-Region Algorithms to First- and Second-Order Critical Points
• Derivative-Free Optimization of Expensive Functions with Computational Error Using Weighted Regression
• A Simplex Method for Function Minimization
• Wedge trust region method for derivative free optimization.

This page was built for publication: A Derivative-Free Trust-Region Algorithm for the Optimization of Functions Smoothed via Gaussian Convolution Using Adaptive Multiple Importance Sampling