Unified Algorithm to Improve Reinforcement Learning in Dynamic Environments - An Instance-based Approach.
Richardson RibeiroFábio FavarimMarco A. C. BarbosaAndré Pinz BorgesOsmar Betazzi DordalAlessandro L. KoerichFabrício EnembreckPublished in: ICEIS (1) (2012)
Keyphrases
- dynamic environments
- reinforcement learning
- learning algorithm
- dynamic programming
- path planning
- preprocessing
- search space
- worst case
- model free
- segmentation algorithm
- objective function
- detection algorithm
- intelligent agents
- collision avoidance
- single agent
- evaluation function
- stochastic approximation
- autonomous agents
- virtual environment
- expectation maximization
- state space
- mobile robot