Reinforcement learning with optimized reward function for stealth applications.
Matheus R. F. MendonçaHeder S. BernardinoRaul Fonseca NetoPublished in: Entertain. Comput. (2018)
Keyphrases
- reward function
- reinforcement learning
- reinforcement learning algorithms
- markov decision processes
- state space
- optimal policy
- policy search
- partially observable
- inverse reinforcement learning
- multiple agents
- transition model
- markov decision process
- function approximation
- model free
- hierarchical reinforcement learning
- state variables
- markov decision problems
- action selection
- transition probabilities
- dynamic programming
- learning agent
- machine learning
- state abstraction
- policy iteration
- function approximators
- average reward
- state action
- approximate dynamic programming
- prior knowledge
- decision makers