OCTOPUS: Deep Reinforcement Learning for Holistic Smart Building Control.
Xianzhong DingWan DuAlberto CerpaPublished in: BuildSys@SenSys (2019)
Keyphrases
- reinforcement learning
- control problems
- optimal control
- control system
- robot control
- control policy
- dynamic programming
- reinforcement learning algorithms
- machine learning
- function approximation
- control method
- adaptive control
- learning process
- control strategies
- neural network
- optimal policy
- learning problems
- model free
- action selection
- artificial neural networks