Linearly convergent bilevel optimization with single-step inner methods
From MaRDI portal
Publication:6155063
Abstract: We propose a new approach to solving bilevel optimization problems, intermediate between solving full-system optimality conditions with a Newton-type approach, and treating the inner problem as an implicit function. The overall idea is to solve the full-system optimality conditions, but to precondition them to alternate between taking steps of simple conventional methods for the inner problem, the adjoint equation, and the outer problem. While the inner objective has to be smooth, the outer objective may be nonsmooth subject to a prox-contractivity condition. We prove linear convergence of the approach for combinations of gradient descent and forward-backward splitting with exact and inexact solution of the adjoint equation. We demonstrate good performance on learning the regularization parameter for anisotropic total variation image denoising, and the convolution kernel for image deconvolution.
Recommendations
- A first order method for solving convex bilevel optimization problems
- Methodology and first-order algorithms for solving nonsmooth and non-strongly convex bilevel optimization problems
- Optimality Conditions for Bilevel Imaging Learning Problems with Total Variation Regularization
- Bilevel optimization with nonsmooth lower level problems
- An inertial extrapolation method for convex simple bilevel optimization
Cites work
- scientific article; zbMATH DE number 46303 (Why is no real title available?)
- scientific article; zbMATH DE number 2155014 (Why is no real title available?)
- A bilevel approach for parameter learning in inverse problems
- A bilevel optimization approach for parameter learning in variational models
- A first order method for solving convex bilevel optimization problems
- A first-order primal-dual algorithm for convex problems with applications to imaging
- A primal-dual hybrid gradient method for nonlinear operators with applications to MRI
- Acceleration and global convergence of a first-order primal-dual method for nonconvex problems
- An inertial extrapolation method for convex simple bilevel optimization
- Application of particle swarm optimization based on CHKS smoothing function for solving nonlinear bilevel programming problem
- Bi-CGSTAB: A Fast and Smoothly Converging Variant of Bi-CG for the Solution of Nonsymmetric Linear Systems
- Bilevel optimization for calibrating point spread functions in blind deconvolution
- Bilevel parameter learning for higher-order total variation regularisation models
- Bilevel programming problems. Theory, algorithms and applications to energy networks
- Convergence analysis of primal-dual algorithms for a saddle-point problem: from contraction perspective
- Directional necessary optimality conditions for bilevel programs
- Dynamic sampling schemes for optimal noise learning under multiple nonsmooth constraints
- Gauss-Newton-type methods for bilevel optimization
- Generalized Nash equilibrium problems, bilevel programming and MPEC. Based on lectures given at the international center for pure and applied mathematics (CIMPA) school, Delhi, India, November 25 -- December 6, 2013
- Image denoising: learning the noise model via nonsmooth PDE-constrained optimization
- Inexact derivative-free optimization for bilevel learning
- Learning consistent discretizations of the total variation
- Mathematical Programs with Equilibrium Constraints
- On bilevel programming. I: General nonlinear cases
- Optimal selection of the regularization function in a weighted total variation model. II: Algorithm, its analysis and numerical tests
- Optimality Conditions for Bilevel Imaging Learning Problems with Total Variation Regularization
- Optimality conditions for bilevel programming problems
- Solving bilevel programs with the KKT-approach
- Solving ill-posed bilevel programs
- Sufficient optimality conditions in bilevel programming
- Techniques for gradient-based bilevel optimization with non-smooth lower level problems
- Testing and non-linear preconditioning of the proximal point method
- The bilevel programming problem: reformulations, constraint qualifications and optimality conditions
- The structure of optimal parameters for image restoration problems
This page was built for publication: Linearly convergent bilevel optimization with single-step inner methods
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6155063)
