Optimal single ordering policy with multiple delivery modes and Bayesian information updates · Study