Scaling Limits of Wide Neural Networks with Weight Sharing: Gaussian Process Behavior, Gradient Independence, and Neural Tangent Kernel Derivation.
Greg YangPublished in: CoRR (2019)
Keyphrases
- gaussian process
- neural network
- gaussian processes
- gaussian process regression
- reproducing kernel hilbert space
- covariance function
- approximate inference
- kernel logistic regression
- regression model
- model selection
- hyperparameters
- gaussian process classification
- latent variables
- artificial neural networks
- sparse approximations
- bayesian framework
- back propagation
- semi supervised
- gaussian process models
- expectation propagation
- human pose estimation
- support vector
- error rate
- edge detection
- kernel methods
- missing values
- cross validation
- synaptic weights