thompson sampling

Thompson sampling is a Bayesian approach to sequential decision-making and reinforcement learning that balances exploration and exploitation by selecting actions according to their probability of being optimal based on current knowledge.

13 papers