Maximum principle based algorithms for deep learning
From MaRDI portal
Recommendations
- Dynamical Systems andOptimal Control Approach to Deep Learning
- A mean-field optimal control formulation of deep learning
- Deep learning as optimal control problems: models and numerical methods
- Substantiation of the backpropagation technique via the Hamilton-Pontryagin formalism for training nonconvex nonsmooth neural networks
- Selection dynamics for deep neural networks
Cites work
- scientific article; zbMATH DE number 5813526 (Why is no real title available?)
- scientific article; zbMATH DE number 1818892 (Why is no real title available?)
- scientific article; zbMATH DE number 1095138 (Why is no real title available?)
- scientific article; zbMATH DE number 3380106 (Why is no real title available?)
- A Maximum Principle of the Pontryagin Type for Systems Described by Nonlinear Difference Equations
- A Stochastic Approximation Method
- A proposal on machine learning via dynamical systems
- Adaptive subgradient methods for online learning and stochastic optimization
- Asymptotic Properties of Non-Linear Least Squares Estimators
- Calculus of variations and optimal control theory. A concise introduction
- Deep learning
- Gradient Theory of Optimal Flight Paths
- Introduction to the mathematical theory of control
- Large-scale machine learning with stochastic gradient descent
- Learning deep architectures for AI
- Method of successive approximations for solution of optimal control problems
- Modifications of the method of successive approximations for solving optimal control problems
- Multiplier and gradient methods
- Nonlinear Programming
- Numerical investigation of a class of Liouville control problems
- On the limited memory BFGS method for large scale optimization
- Optimal control of continuity equations
- Stable architectures for deep neural networks
- Survey of Numerical Methods for Trajectory Optimization
- The Fokker-Planck equation. Methods of solutions and applications.
- The discrete-time maximum principle : a survey and some new results
- The maximum principle in optimal control, then and now
Cited in
(67)- Variational networks: an optimal control approach to early stopping variational methods for image restoration
- A mean-field optimal control formulation of deep learning
- Cluster-based classification with neural ODEs via control
- A measure theoretical approach to the mean-field maximum principle for training NeurODEs
- Machine learning from a continuous viewpoint. I
- Principled deep neural network training through linear programming
- A mathematical perspective on transformers
- An optimal control framework for adaptive neural ODEs
- Deep learning approximation of diffeomorphisms via linear-control systems
- Sparsity in long-time control of neural ODEs
- Control on the manifolds of mappings with a view to the deep learning
- Do ideas have shape? Idea registration as the continuous limit of artificial neural networks
- Machine learning and computational mathematics
- A Max-Sum algorithm for training discrete neural networks
- Deep limits of residual neural networks
- scientific article; zbMATH DE number 7625195 (Why is no real title available?)
- A modified MSA for stochastic control problems
- Selection dynamics for deep neural networks
- Deep neural networks motivated by partial differential equations
- Detecting the most probable transition pathway based on optimal control theory
- Linear Convergence of a Policy Gradient Method for Some Finite Horizon Continuous Time Control Problems
- Solving optimal control problems of rigid-body dynamics with collisions using the hybrid minimum principle
- Turnpike in optimal control of PDEs, ResNets, and beyond
- The Mori-Zwanzig formulation of deep learning
- Layer-parallel training of residual networks with auxiliary variable networks
- From NeurODEs to AutoencODEs: a mean-field control framework for width-varying neural networks
- Maximal Function Pooling with Applications
- Deep learning as optimal control problems: models and numerical methods
- Interpolation, approximation, and controllability of deep neural networks
- Approximation capabilities of measure-preserving neural networks
- An artificial neural network approach to identify the parameter in a nonlinear subdiffusion model
- Personalized algorithm generation: a case study in learning ODE integrators
- scientific article; zbMATH DE number 7387622 (Why is no real title available?)
- Deep learning via dynamical systems: an approximation perspective
- On the convergence of conjugate gradient and GMRES algorithms in the forward backward sweep method for optimal control
- Gradient convergence of deep learning-based numerical methods for BSDEs
- Optimal control using to approximate probability distribution of observation set
- On dissipativity of cross-entropy loss in training ResNets. A turnpike towards architecture search
- On mathematical modeling in image reconstruction and beyond
- On the regularized risk of distributionally robust learning over deep neural networks
- Neural ODEs as the deep limit of ResNets with constant weights
- Explicit Stabilized Integrators for Stiff Optimal Control Problems
- Synchronizing pretrained kernel regressors with applications to American option pricing
- A modified method of successive approximations for stochastic recursive optimal control problems
- Algorithms for solving high dimensional PDEs: from nonlinear Monte Carlo to machine learning
- Substantiation of the backpropagation technique via the Hamilton-Pontryagin formalism for training nonconvex nonsmooth neural networks
- Dual non-autonomous deep convolutional neural network for image denoising
- Learning symmetries and non-Euclidean data representations via collective dynamics of generalized Kuramoto oscillators
- Concurrent learning for adaptive Pontryagin's maximum principle of nonlinear systems with inequality constraints
- A sequential quadratic Hamiltonian algorithm for training explicit RK neural networks
- Dynamical Systems andOptimal Control Approach to Deep Learning
- Symplectic Runge-Kutta discretization of a regularized forward-backward sweep iteration for optimal control problems
- Efficient and stable SAV-based methods for gradient flows arising from deep learning
- A review on deep learning in medical image reconstruction
- A closed-loop symplectic regularized algorithm for constrained time-varying optimal control
- Predict globally, correct locally: parallel-in-time optimization of neural networks
- Mirror descent for stochastic control problems with measure-valued controls
- The Pontryagin maximum principle for training convolutional neural networks
- Estimating adsorption isotherm parameters in chromatography via a virtual injection promoting double feed-forward neural network
- Sub-Riemannian landmark matching and its interpretation as residual neural networks
- Structure-preserving deep learning
- On the universal approximation property of deep fully convolutional neural networks
- An optimal control method to compute the most likely transition path for stochastic dynamical systems with jumps
- Artificial neural network approximations of Cauchy inverse problem for linear PDEs
- A mathematical perspective of machine learning
- Towards a mathematical understanding of neural network-based machine learning: what we know and what we don't
- On the Banach spaces associated with multi-layer ReLU networks: function representation, approximation theory and gradient descent dynamics
Describes a project that uses
Uses Software
This page was built for publication: Maximum principle based algorithms for deep learning
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4558489)
