Self-Activating Neural Ensembles for Continual Reinforcement Learning.
Sam PowersEliot XingAbhinav GuptaPublished in: CoLLAs (2022)
Keyphrases
- reinforcement learning
- fitted q iteration
- function approximation
- network architecture
- neural network
- state space
- sensory inputs
- artificial neural
- ensemble methods
- machine learning
- neural model
- nonlinear predictive control
- data sets
- temporal difference
- bio inspired
- ensemble learning
- optimal control
- learning problems
- optimal policy
- biologically inspired
- decision trees
- multi agent
- robotic control
- ensemble selection
- reinforcement learning methods
- temporal difference learning
- reinforcement learning algorithms
- markov decision processes
- associative memory
- classifier ensemble
- supervised learning
- artificial neural networks
- action selection
- neural computation
- model free
- base classifiers