Accelerating the Deep Reinforcement Learning with Neural Network Compression.
Hongjie ZhangZhuocheng HeJing LiPublished in: IJCNN (2019)
Keyphrases
- neural network
- reinforcement learning
- learning capabilities
- back propagation
- function approximation
- data compression
- artificial neural networks
- function approximators
- image compression
- reinforcement learning algorithms
- compression ratio
- pattern recognition
- optimal policy
- feed forward neural networks
- markov decision processes
- compression algorithm
- compression scheme
- multi layer perceptron
- neural network model
- state space
- neural network is trained
- supervised learning
- recurrent neural networks
- learning algorithm
- optimal control
- multi layer
- compression rate
- robot control
- lossless compression
- markov decision process
- action space
- network model
- neural nets
- auto associative
- fault diagnosis
- genetic algorithm
- machine learning
- fuzzy artmap
- policy search
- robotic control
- model free
- bp neural network
- self organizing maps
- image quality
- expert systems
- multi agent