Optimal Regret of Bandits under Differential Privacy | Read Paper on Bytez