KL-UCB-Based Policy for Budgeted Multi-Armed Bandits with Stochastic Action Costs.
Ryo WatanabeJunpei KomiyamaAtsuyoshi NakamuraMineichi KudoPublished in: IEICE Trans. Fundam. Electron. Commun. Comput. Sci. (2017)
Keyphrases
- multi armed bandits
- multi armed bandit
- bandit problems
- reinforcement learning
- action selection
- expected cost
- optimal policy
- decision problems
- action space
- total cost
- optimal strategy
- markov decision problems
- regret bounds
- multi class
- average cost
- function approximation
- long run
- infinite horizon
- decision processes
- state space