An optimal control framework for adaptive neural ODEs

DOI10.1007/S10444-024-10149-0MaRDI QIDQ6561374zbMATH OpenFDO

Authors Joubine Aghili, O. Mula

Publication date 25 June 2024

Published in Advances in Computational Mathematics (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/S10444-024-10149-0

zbMATH Keywords

adaptivitydeep learningneural ODE networksoptimal control

Mathematics Subject Classification ID

Existence theories for optimal control problems involving ordinary differential equations (49J15) Numerical optimization and variational techniques (65K10) Numerical solution of boundary value problems involving ordinary differential equations (65L10) Computational learning theory (68Q32) Learning and adaptive systems in artificial intelligence (68T05)

Recommendations

• Deep learning as optimal control problems: models and numerical methods
• Neural ODE Control for Classification, Approximation, and Transport
• Sparsity in long-time control of neural ODEs
• A measure theoretical approach to the mean-field maximum principle for training NeurODEs
• Neural ODEs as the deep limit of ResNets with constant weights

Cites work

• A ``parareal in time discretization of PDE's
• A mean-field optimal control formulation of deep learning
• A proposal on machine learning via dynamical systems
• An adaptive parareal algorithm
• Deep learning as optimal control problems: models and numerical methods
• Deep learning via dynamical systems: an approximation perspective
• Deep neural networks, generic universal interpolation, and controlled ODEs
• Error bounds for approximations with deep ReLU networks
• Error bounds for approximations with deep ReLU neural networks in \(W^{s , p}\) norms
• scientific article; zbMATH DE number 1095138 (Why is no real title available?)
• Introduction to the mathematical theory of control
• Layer-Parallel Training of Deep Residual Neural Networks
• Maximum principle based algorithms for deep learning
• Method of successive approximations for solution of optimal control problems
• Neural architecture search: a survey
• Neural ODE Control for Classification, Approximation, and Transport
• On shooting methods for boundary value problems
• Optimal order a posteriori error estimates for a class of Runge-Kutta and Galerkin methods
• Remarks on Inequalities for Large Deviation Probabilities
• Stable architectures for deep neural networks
• The maximum principle in optimal control, then and now

Cited in

(4)

• Sparsity in long-time control of neural ODEs
• Enhancing continuous time series modelling with a latent ODE-LSTM approach
• Implementation and (inverse modified) error analysis for implicitly templated ODE-nets
• Towards continuous mathematical models for the analysis of classes of deep neural networks

This page was built for publication: An optimal control framework for adaptive neural ODEs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6561374)