Influence Diagram Bandits: Variational Thompson Sampling for Structured Bandit Problems.
Tong YuBranislav KvetonZheng WenRuiyi ZhangOle J. MengshoelPublished in: CoRR (2020)
Keyphrases
- bandit problems
- decision problems
- influence diagrams
- multi armed bandits
- exploration exploitation
- multi armed bandit problems
- decision analysis
- expected utility
- limited memory
- utility function
- decision making under uncertainty
- optimal policy
- multi armed bandit
- optimal strategy
- image segmentation
- computational complexity
- bayesian decision problems
- np hard
- monte carlo
- graphical representation
- stochastic dynamic programming
- probabilistic inference
- sensitivity analysis
- bayesian networks
- extensive form games
- decision making
- machine learning
- graphical models
- sample size
- markov chain