We propose the first learning algorithm for single-product, periodic-review, backlogging inventory systems with random … cyclic stochastic gradient descent type of algorithm whose running average cost asymptotically converges to the clairvoyant … optimal cost. We prove that the rate of convergence guarantee of our algorithm is $O(1/\sqrt{T})$, which is provably tight for …

propose a non-parametric algorithm that generates a sequence of adaptive ordering decisions based on the stochastic gradient … descent method. We compare the T-period cost of our algorithm to the clairvoyant, who knows the underlying demand distribution …