Bayesian Optimization -- Multi-Armed Bandit Problem