j2kun / exp3

Python code for the post "Adversarial Bandits and the Exp3 Algorithm"

exp3

Python code for the post "Adversarial Bandits and the Exp3 Algorithm"

exp3's Issues

Enable Sourcegraph

I want to use Sourcegraph for exp3 code search, browsing, and usage examples. Can an admin enable Sourcegraph for this repository? Just go to https://sourcegraph.com/github.com/j2kun/exp3. (It should only take 30 seconds.)

Thank you!

if numRounds is unknown, how should bestUpperBoundEstimate and gamma be set?

1.Hello, I wonder that in the exp3 algorithm, if numRounds is unknown, how should bestUpperBoundEstimate and gamma be set?(I saw that "You can try both: https://github.com/j2kun/exp3/blob/master/exp3.py#L39-L40")

2.In the algorithm, the rewards of each round are counted through the number of rounds, and the weight of the new round is calculated. In the business, can we use days as the unit to count the daily rewards and calculate the weight of the new day？If it could work, how should bestUpperBoundEstimate and gamma be set?