A genetic fuzzy system for interpretable and parsimonious reinforcement learning policies.
Jordan T. BishopMarcus GallagherWill N. BrownePublished in: GECCO Companion (2021)
Keyphrases
- reinforcement learning
- optimal policy
- policy search
- markov decision process
- control policies
- markov decision processes
- reward function
- reinforcement learning agents
- markov decision problems
- fitted q iteration
- state space
- partially observable markov decision processes
- total reward
- cooperative multi agent systems
- policy gradient methods
- function approximation
- learning algorithm
- macro actions
- dynamic programming
- temporal difference
- continuous state
- control policy
- decision problems
- hierarchical reinforcement learning
- reinforcement learning algorithms
- multi agent
- function approximators
- action selection
- natural actor critic
- decision processes
- policy iteration
- model free
- long run
- approximate policy iteration
- neural network
- state and action spaces
- supervised learning
- multiagent reinforcement learning
- transfer learning
- learning problems
- temporal difference learning
- action space
- optimal control
- robot control