Fast and slow curiosity for high-level exploration in reinforcement learning.
Nicolas BougieRyutaro IchisePublished in: Appl. Intell. (2021)
Keyphrases
- reinforcement learning
- high level
- low level
- active exploration
- exploration strategy
- exploration exploitation
- action selection
- model based reinforcement learning
- reinforcement learning algorithms
- function approximation
- autonomous learning
- machine learning
- exploration exploitation tradeoff
- active learning
- semantic information
- low level features
- transfer learning
- higher level
- optimal policy
- programming language
- state space
- markov decision processes
- balancing exploration and exploitation
- learning algorithm
- model free
- data sets
- neural network
- search engine
- decision making
- case study
- function approximators
- markov decision process
- mid level
- objective function
- multi agent
- lower level
- optimal control
- learning process
- dynamic programming
- search strategies
- real time