On the ISS Property of the Gradient Flow for Single Hidden-Layer Neural Networks with Linear Activations.
Arthur Castello B. de OliveiraMilad SiamiEduardo D. SontagPublished in: CoRR (2023)
Keyphrases
- neural network
- single hidden layer
- feedforward neural networks
- activation function
- gradient flow
- extreme learning machine
- hidden layer
- feed forward neural networks
- back propagation
- pattern recognition
- recurrent neural networks
- hidden neurons
- fuzzy logic
- multilayer perceptron
- neural network model
- active contour model
- active contours
- artificial neural networks
- energy functional
- level set
- computer vision
- partial differential equations
- neural nets
- fuzzy neural network
- hidden nodes
- curve evolution
- multiscale
- genetic algorithm