Model-free POMDP optimisation of tutoring systems with echo-state networks.
Lucie DaubigneyMatthieu GeistOlivier PietquinPublished in: SIGDIAL Conference (2013)
Keyphrases
- model free
- tutoring system
- echo state networks
- reinforcement learning
- policy evaluation
- recurrent neural networks
- function approximation
- reinforcement learning algorithms
- temporal difference
- student modelling
- intelligent tutoring systems
- partially observable markov decision processes
- state space
- policy iteration
- partially observable
- markov decision processes
- average reward
- optimal policy
- multi agent
- genetic algorithm
- neural network
- learning algorithm
- cognitive skills
- least squares
- markov decision process
- authoring tool
- function approximators
- finite state
- optimal control
- feed forward
- dynamical systems
- transfer learning
- learning process
- machine learning