Pre-training Neural Networks with Human Demonstrations for Deep Reinforcement Learning.
Gabriel Victor de la CruzYunshu DuMatthew E. TaylorPublished in: CoRR (2017)
Keyphrases
- neural network
- reinforcement learning
- training process
- training algorithm
- back propagation
- pattern recognition
- backpropagation algorithm
- feedforward neural networks
- supervised learning
- neural network training
- feed forward neural networks
- human subjects
- learning algorithm
- multi layer perceptron
- learning capabilities
- function approximation
- human behavior
- neural network model
- deep architectures
- artificial neural networks
- training samples
- human experts
- fuzzy logic
- machine learning
- genetic algorithm
- test set
- fuzzy systems
- recurrent networks
- online learning
- function approximators
- learning process
- reinforcement learning algorithms
- neural network structure
- radial basis function network
- state space
- temporal difference
- training examples
- transfer learning
- markov decision processes
- evaluation function
- learning problems