Validation of a Reinforcement Learning Policy for Dosage Optimization of Erythropoietin.
José David Martín-GuerreroEmilio Soria-OlivasMarcelino Martínez-SoberMónica Climente-MartíTeresa De Diego-SantosN. Víctor JiménezPublished in: Australian Conference on Artificial Intelligence (2007)
Keyphrases
- reinforcement learning
- optimal policy
- policy search
- action selection
- optimization problems
- global optimization
- state space
- function approximation
- markov decision processes
- markov decision process
- optimization process
- partially observable environments
- transition model
- policy gradient
- action space
- optimization methods
- optimization algorithm
- multi agent
- temporal difference
- reward function
- partially observable
- combinatorial optimization
- control policies
- continuous state
- sufficient conditions
- policy evaluation
- evolutionary algorithm
- reinforcement learning problems
- objective function