Comparison of On-Policy Deep Reinforcement Learning A2C with Off-Policy DQN in Irrigation Optimization: A Case Study at a Site in Portugal.
Khadijeh AlibabaeiPedro Dinis GasparEduardo AssunçãoSaeid AlirezazadehTânia M. LimaVasco N. G. J. SoaresJoão M. L. P. CaldeiraPublished in: Comput. (2022)
Keyphrases
- reinforcement learning
- optimal policy
- global optimization
- optimization process
- function approximation
- control policy
- website
- action selection
- policy search
- optimization algorithm
- partially observable environments
- markov decision process
- case study
- markov decision processes
- test bed
- state space
- model free
- control policies
- state action
- function approximators
- optimization method
- united kingdom
- rl algorithms
- average reward
- partially observable
- lecture notes in artificial intelligence
- dynamic programming
- optimization problems
- partially observable domains
- agent receives
- policy gradient methods
- state and action spaces
- markov decision problems
- reward function
- learning problems
- learning process