Stochastic approximation for discrete optimization of noisy loss measurements

摘要

We consider the stochastic optimization of a noisy convex loss function defined on p-dimensional grid of points in Euclidean space. We introduce the middle point discrete simultaneous perturbation stochastic approximation (DSPSA) algorithm to this discrete space. Consistent with other stochastic approximation methods, this method formally accommodates noisy measurements of the loss function.