Bayesian Bandit algorithm
Bayesian Bandit algorithm
(estimate-value {:keys [pulls value prior] :as arm})
(reward arm reward)
(select-arm arms)
Returns the arm that should be pulled. Arm records should have a :prior value (assumed to be 1.0 if no prior set).
Returns the arm that should be pulled. Arm records should have a :prior value (assumed to be 1.0 if no prior set).
cljdoc is a website building & hosting documentation for Clojure/Script libraries
× close