Participants
Qi Wang
Discrete Simultaneous Perturbation Stochastic Approximation On Loss Function with Noisy Measurements
We consider the stochastic optimization of a loss function defined on p-dimensional grid of points in Euclidean space. We introduce the middle point discrete simultaneous perturbation stochastic approximation (DSPSA) algorithm for such discrete problems, and show that convergence to the minimum is achieved. Consistent with other stochastic approximation methods, this method formally accommodates noisy measurements of the loss function. We also consider the rate of convergence of this method.
Rate of Convergence Analysis of Model-Free Algorithm
Model-free adaptive control of nonlinear stochastic systems is widely used in many areas. Spall and Cristion (1998) uses the simultaneous perturbation stochastic approximation (SPSA) algorithm to estimate the controller of a nonlinear stochastic system with unknown dynamics. We consider the rate of convergence analysis of SPSA in the model-free control.