On Developing a UAV Pursuit-Evasion Policy Using Reinforcement Learning.
Bogdan I. VlahovEric SquiresLaura StricklandCharles PippinPublished in: ICMLA (2018)
Keyphrases
- pursuit evasion
- reinforcement learning
- optimal policy
- policy search
- action selection
- markov decision process
- partially observable
- function approximation
- partially observable environments
- policy evaluation
- function approximators
- control policy
- actor critic
- state action
- state and action spaces
- learning algorithm
- policy iteration
- vision based navigation
- state space
- markov decision processes
- model free
- decision problems
- control problems
- dynamic environments
- approximate dynamic programming
- policy gradient
- control policies
- reinforcement learning algorithms
- rl algorithms
- markov decision problems
- continuous state
- policy making
- control algorithm
- reward function
- partially observable domains
- transition model
- action space
- partially observable markov decision processes
- unmanned aerial vehicles
- agent learns
- reinforcement learning problems
- average reward
- path planning
- decision making