Memory-Efficient Episodic Control Reinforcement Learning with Dynamic Online k-means.
Andrea AgostinelliKai ArulkumaranMarta SarricoPierre RichemondAnil Anthony BharathPublished in: CoRR (2019)
Keyphrases
- memory efficient
- reinforcement learning
- k means
- control problems
- online learning
- hierarchical clustering
- integral image
- external memory
- robot control
- function approximation
- control method
- markov decision processes
- dynamic environments
- optimal control
- action selection
- state space
- multi agent
- clustering method
- spectral clustering
- machine learning
- control strategy
- adaptive control
- expectation maximization
- control system
- inverted pendulum
- clustering algorithm