Universal Function Approximation by Deep Neural Nets with Bounded Width and ReLU Activations.
Boris HaninPublished in: CoRR (2017)
Keyphrases
- function approximation
- neural nets
- learning tasks
- reinforcement learning
- feed forward
- temporal difference learning algorithms
- neural network
- radial basis function
- back propagation
- temporal difference learning
- artificial neural networks
- temporal difference
- machine learning
- supervised learning
- rbfnn
- hidden layer
- function approximators
- learning experience
- transfer learning
- nearest neighbor
- genetic algorithm
- learning algorithm
- temporal difference methods
- machine learning algorithms