HARL: Hierarchical Adaptive Reinforcement Learning Based Auto Scheduler for Neural Networks.
Zining ZhangBingsheng HeZhenjie ZhangPublished in: ICPP (2022)
Keyphrases
- neural network
- reinforcement learning
- learning capabilities
- function approximation
- pattern recognition
- function approximators
- genetic algorithm
- artificial neural networks
- fuzzy logic
- fault diagnosis
- hierarchical structure
- reinforcement learning algorithms
- model free
- multi layer
- adaptive control
- coarse to fine
- rule extraction
- back propagation
- state space
- multiscale
- electric vehicles
- scheduling algorithm
- evaluation function
- neural network model
- markov decision processes
- data driven
- machine learning