The idea behind Thompson Sampling is based on keeping a posterior distribution
The idea behind Thompson Sampling is based on keeping a posterior distribution
 over functions in some family  after observing the first
 over functions in some family  after observing the first