In a multi-armed bandit, the goal is to maximize the expected cumulative reward. This goal is usually (equivalently?) Expressed in terms of expected cumulative regret.
Question: Why not just face the reward? Why formulate the goal in terms of regret?
T-mount from one side and something strange (similar to Arri PL? From the other)?
There was a difference between events and achievement of goals in Google Analytics.
These events are sent from Salesforce using the measurement protocol and do not match the achievement of the objectives.
Do you have any suggestions or ideas?
Thank you in advance for your time.