تقرير
Generalized Deterministic Perturbations For Stochastic Gradient Search
| العنوان: | Generalized Deterministic Perturbations For Stochastic Gradient Search |
|---|---|
| المؤلفون: | Chandramouli, K., Prabuchandran, K. J., Reddy, D. Sai Koti, Bhatnagar, Shalabh |
| سنة النشر: | 2017 |
| المجموعة: | Computer Science |
| مصطلحات موضوعية: | Computer Science - Systems and Control |
| الوصف: | Stochastic optimization (SO) considers the problem of optimizing an objective function in the presence of noise. Most of the solution techniques in SO estimate gradients from the noise corrupted observations of the objective and adjust parameters of the objective along the direction of the estimated gradients to obtain locally optimal solutions. Two prominent algorithms in SO namely Random Direction Kiefer-Wolfowitz (RDKW) and Simultaneous Perturbation Stochastic Approximation (SPSA) obtain noisy gradient estimate by randomly perturbing all the parameters simultaneously. This forces the search direction to be random in these algorithms and causes them to suffer additional noise on top of the noise incurred from the samples of the objective. Owing to this additional noise, the idea of using deterministic perturbations instead of random perturbations for gradient estimation has also been studied. Two specific constructions of the deterministic perturbation sequence using lexicographical ordering and Hadamard matrices have been explored and encouraging results have been reported in the literature. In this paper, we characterize the class of deterministic perturbation sequences that can be utilized in the RDKW algorithm. This class expands the set of known deterministic perturbation sequences available in the literature. Using our characterization we propose a construction of a deterministic perturbation sequence that has the least possible cycle length among all deterministic perturbations. Through simulations we illustrate the performance gain of the proposed deterministic perturbation sequence in the RDKW algorithm over the Hadamard and the random perturbation counterparts. We establish the convergence of the RDKW algorithm for the generalized class of deterministic perturbations. Comment: Accepted in Control and Decision Conference |
| نوع الوثيقة: | Working Paper |
| URL الوصول: | http://arxiv.org/abs/1702.06250 |
| رقم الأكسشن: | edsarx.1702.06250 |
| قاعدة البيانات: | arXiv |
| الوصف غير متاح. |