Learning continuous-time working memory tasks with on-policy neural reinforcement learning.
Davide ZambranoPieter R. RoelfsemaSander M. BohtéPublished in: Neurocomputing (2021)
Keyphrases
- reinforcement learning
- learning process
- cognitive load
- learning algorithm
- working memory
- state space
- actor critic
- optimal control
- neural network
- computer vision
- optimal policy
- computational model
- function approximators
- action selection
- fitted q iteration
- prefrontal cortex
- cognitive architecture
- function approximation
- learning tasks
- knowledge acquisition
- database systems
- multimedia