Stateful Posted Pricing with Vanishing Regret via Dynamic Deterministic Markov Decision Processes.
Yuval EmekRon LaviRad NiazadehYangguang ShiPublished in: NeurIPS (2020)
Keyphrases
- markov decision processes
- reward function
- finite state
- state space
- total reward
- reinforcement learning
- dynamic programming
- optimal policy
- transition matrices
- reinforcement learning algorithms
- risk sensitive
- stationary policies
- policy iteration
- finite horizon
- decision theoretic planning
- decision processes
- planning under uncertainty
- infinite horizon
- partially observable
- markov decision process
- average reward
- reachability analysis
- average cost
- action space
- factored mdps
- expected reward
- model based reinforcement learning
- state abstraction
- dynamic environments
- stochastic shortest path
- real time dynamic programming
- multistage
- lower bound