Active neuron least squares: a training method for multivariate rectified neural networks
From MaRDI portal
(Redirected from Publication:5095494)
Recommendations
- Adaptive two-layer ReLU neural network. I: Best least-squares approximation
- Training recurrent neural networks by sequential least squares and the alternating direction method of multipliers
- Variable projections neural network training
- Active learning in neural networks
- scientific article; zbMATH DE number 1843080
Cites work
- A Family of Variable-Metric Methods Derived by Variational Means
- A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems
- A new approach to variable metric algorithms
- Adaptive activation functions accelerate convergence in deep and physics-informed neural networks
- Adaptive restart for accelerated gradient schemes
- Conditioning of Quasi-Newton Methods for Function Minimization
- DeepONet
- Dying ReLU and initialization: theory and numerical examples
- Efficient implementation of weighted ENO schemes
- Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations
- scientific article; zbMATH DE number 3850830 (Why is no real title available?)
- On the eigenvector bias of Fourier feature networks: from regression to solving multi-scale PDEs with physics-informed neural networks
- Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- Plateau phenomenon in gradient descent training of RELU networks: explanation, quantification, and avoidance
- Solving high-dimensional partial differential equations using deep learning
- The Convergence of a Class of Double-rank Minimization Algorithms 1. General Considerations
- Understanding and Mitigating Gradient Flow Pathologies in Physics-Informed Neural Networks
Cited in
(10)- Least-squares ReLU neural network (LSNN) method for scalar nonlinear hyperbolic conservation law
- Greedy training algorithms for neural networks and applications to PDEs
- Locally adaptive activation functions with slope recovery for deep and physics-informed neural networks
- Training recurrent neural networks by sequential least squares and the alternating direction method of multipliers
- On the training and generalization of deep operator networks
- Subspace method based on neural networks for solving the partial differential equation in weak form
- Efficient shallow Ritz method for 1D diffusion problems
- Randomized greedy algorithms for neural network optimization in solving partial differential equations
- Multiscale neural networks for approximating Green's functions
- Common pitfalls to avoid while using multiobjective optimization in machine learning
This page was built for publication: Active neuron least squares: a training method for multivariate rectified neural networks
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5095494)
