Data-Driven Upper Confidence Bounds with Near-Optimal Regret for Heavy-Tailed Bandits.

Ambrus Tams, Szabolcs Szentpteri, Balzs Csand Csji

Browse the full AISTATS paper archive.