An outer-approximation guided optimization approach for constrained neural network inverse problems
From MaRDI portal
Abstract: This paper discusses an outer-approximation guided optimization method for constrained neural network inverse problems with rectified linear units. The constrained neural network inverse problems refer to an optimization problem to find the best set of input values of a given trained neural network in order to produce a predefined desired output in presence of constraints on input values. This paper analyzes the characteristics of optimal solutions of neural network inverse problems with rectified activation units and proposes an outer-approximation algorithm by exploiting their characteristics. The proposed outer-approximation guided optimization comprises primal and dual phases. The primal phase incorporates neighbor curvatures with neighbor outer-approximations to expedite the process. The dual phase identifies and utilizes the structure of local convex regions to improve the convergence to a local optimal solution. At last, computation experiments demonstrate the superiority of the proposed algorithm compared to a projected gradient method.
Recommendations
- Design and analysis of an efficient neural network model for solving nonlinear optimization problems
- Artificial Intelligence and Soft Computing - ICAISC 2004
- Neural network for constrained nonsmooth optimization using Tikhonov regularization
- A novel approach based on recurrent neural networks applied to nonlinear systems optimization
- Neuromorphic approach to inverse problems in aerodynamics
Cites work
- scientific article; zbMATH DE number 1818892 (Why is no real title available?)
- An outer-approximation algorithm for a class of mixed-integer nonlinear programs
- Applications of a general convergence theory for outer approximation algorithms
- Cutting planes in integer and mixed integer programming
- Deep neural networks and mixed integer linear optimization
- Elements of Large-Scale Mathematical Programming Part I: Concepts
- Generalized Benders decomposition
- Non-Euclidean restricted memory level method for large-scale convex optimization
- Partitioning procedures for solving mixed-variables programming problems
- The Cutting-Plane Method for Solving Convex Programs
Cited in
(4)- Special issue: Global solution of integer, stochastic and nonconvex optimization problems
- On obtaining sparse semantic solutions for inverse problems, control, and neural network training
- Fixed-point algorithms for inverse of residual rectifier neural networks
- Optimization of neural network training for image recognition based on trigonometric polynomial approximation
This page was built for publication: An outer-approximation guided optimization approach for constrained neural network inverse problems
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2097633)
