Principal-Agent Reward Shaping in MDPs.
Omer Ben-PoratYishay MansourMichal MoshkovitzBoaz TaitlerPublished in: AAAI (2024)
Keyphrases
- reward shaping
- principal agent
- reinforcement learning
- markov decision problems
- markov decision processes
- reinforcement learning algorithms
- reward function
- state space
- partially observable
- markov decision process
- moral hazard
- linear programming
- assembly systems
- optimal policy
- policy search
- policy iteration
- function approximation
- dynamic programming
- complex domains
- utility function
- decision theoretic
- infinite horizon
- learning algorithm
- transition model
- machine learning
- action space
- decision processes
- average cost
- decision making
- finite horizon
- model free
- queueing networks
- continuous state
- transition probabilities
- planning under uncertainty
- finite state