A Recursively Recurrent Neural Network (R2N2) Architecture for Learning Iterative Algorithms
DOI10.1137/22m1535310arXiv2211.12386MaRDI QIDQ6154930
Manuel Dahmen, Ioannis G. Kevrekidis, Yue Guo, Qianxiao Li, Danimir T. Doncevic, Alexander Mitsos, Felix Dietrich
Publication date: 12 March 2024
Published in: SIAM Journal on Scientific Computing (Search for Journal in Brave)
Full work available at URL: https://arxiv.org/abs/2211.12386
Runge-Kutta methodsmachine learningnumerical analysismetalearningNewton-Krylov solversdata-driven algorithm design
Artificial neural networks and deep learning (68T07) Numerical computation of solutions to systems of equations (65H10) Iterative numerical methods for linear systems (65F10) Numerical methods for initial value problems involving ordinary differential equations (65L05) Multistep, Runge-Kutta and extrapolation methods for ordinary differential equations (65L06)
Cites Work
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- How to generate random matrices from the classical compact groups
- On the limited memory BFGS method for large scale optimization
- Jacobian-free Newton-Krylov methods: a survey of approaches and applications.
- Iterative solution of linear systems in the 20th century
- Optimal deterministic algorithm generation
- A polyhedral branch-and-cut approach to global optimization
- Review of nonlinear mixed-integer and disjunctive programming techniques
- Sympnets: intrinsic structure-preserving symplectic networks for identifying Hamiltonian systems
- A machine learning framework for data driven acceleration of computations of differential equations
- A proposal on machine learning via dynamical systems
- SATenstein: automatically building local search SAT solvers from components
- Linear algebra algorithms as dynamical systems
- GMRES: A Generalized Minimal Residual Algorithm for Solving Nonsymmetric Linear Systems
- Solving Nonlinear Equations with Newton's Method
- A variational perspective on accelerated methods in optimization
- Discovering faster matrix multiplication algorithms with reinforcement learning
- Personalized Algorithm Generation: A Case Study in Learning ODE Integrators
- On the Koopman Operator of Algorithms
- A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play
- Mixed-integer nonlinear optimization
- Numerical Methods for Ordinary Differential Equations
- Comparison of MINLP formulations for global superstructure optimization
This page was built for publication: A Recursively Recurrent Neural Network (R2N2) Architecture for Learning Iterative Algorithms
