Reserve a data! O Google I / O retorna de 18 a 20 de maio Registre-se agora

tf_agents.bandits.policies.policy_utilities.PerArmPolicyInfo

PerArmPolicyInfo(log_probability, predicted_rewards_mean, predicted_rewards_optimistic, predicted_rewards_sampled, bandit_policy_type, chosen_arm_features)

log_probability

predicted_rewards_mean

predicted_rewards_optimistic

predicted_rewards_sampled

bandit_policy_type

chosen_arm_features