BO for Contextual Bandits

Posterior Sampling

An algorithmic approach where a sample is drawn from the posterior distribution of the model to guide the next decision, used in contextual bandits and Bayesian optimization.

← Wstecz