Multi-task reinforcement learning with associative memory models considering the multiple distributions of MDPs.
Kinbara FumihiroShohei KatoMunehiro NakamuraPublished in: GCCE (2015)
Keyphrases
- reinforcement learning
- associative memory
- multi task
- multiple tasks
- markov decision processes
- state space
- probabilistic model
- transfer learning
- multi task learning
- sparse learning
- optimal policy
- learning tasks
- gaussian processes
- supervised learning
- model selection
- text categorization
- learning problems
- neural network model
- multitask learning
- kohonen feature map
- learning process
- learning environment