When Do Neural Networks Outperform Kernel Methods?
Behrooz GhorbaniSong MeiTheodor MisiakiewiczAndrea MontanariPublished in: NeurIPS (2020)
Keyphrases
- kernel methods
- neural network
- learning problems
- kernel function
- support vector
- machine learning
- kernel matrix
- feature space
- support vector machine
- pattern recognition
- learning tasks
- artificial neural networks
- similarity function
- reproducing kernel hilbert space
- multiple kernel learning
- back propagation
- kernel learning
- positive semidefinite
- kernel trick
- multiple kernel
- kernel fisher discriminant analysis
- kernel ridge regression
- kernel matrices
- high dimensional feature space
- training process
- training data
- supervised learning