Regarding using Regret as a Loss Function

[SachinVashisth]

Hi

Thanks for a wonderful article related to Multi-Arm Bandits (MAB). In the article, you discussed that the loss function is the total regret we might have by not selecting the optimal action up to the time step T.

In the code, regret is calculated and stored to create a graph when MAB iterations are completed. But how are we using it to make a better selection of arms while MAB is running?

Is the regret only calculated for plotting the graphs OR is it used in some way in the MAB setup?