Implicit bias of gradient descent for mean squared error regression with wide neural networks.
Hui JinGuido MontúfarPublished in: CoRR (2020)
Keyphrases
- neural network
- regression model
- learning rules
- wide range
- artificial neural networks
- neural network model
- pattern recognition
- objective function
- cost function
- regression problems
- back propagation
- neural nets
- regression algorithm
- gaussian processes
- recurrent neural networks
- multilayer perceptron
- classification and regression trees
- support vector regression
- model selection
- loss function
- polynomial regression
- feed forward
- regression methods
- reproducing kernel hilbert space
- activation function
- fuzzy neural network
- regression analysis
- fuzzy systems
- multi layer
- linear regression
- self organizing maps
- hidden layer
- vector quantization
- rule extraction
- quantization error
- fault diagnosis
- genetic programming
- regression function