An adaptive multidimensional version of the Kiefer-Wolfowitz stochastic approximation algorithm

Author

Broadie, Mark ; Cicek, Deniz M. ; Zeevi, Assaf

Author_Institution

Grad. Sch. of Bus., Columbia Univ., New York, NY, USA

fYear

2009

fDate

13-16 Dec. 2009

Firstpage

601

Lastpage

612

Abstract

We extend the scaled-and-shifted Kiefer-Wolfowitz (SSKW) algorithm developed by Broadie, Cicek, and Zeevi (2009) to multiple dimensions. The salient feature of this algorithm is that it makes adjustments of the tuning parameters that adapt to the underlying problem characteristics. We compare the performance of this algorithm to the traditional Kiefer-Wolfowitz (KW) one and observe significant improvement in the finite-time performance on some stylized test functions and a multidimensional newsvendor problem.

Keywords

adaptive systems; approximation theory; multidimensional systems; simulation; stochastic processes; stochastic programming; Kiefer-Wolfowitz stochastic approximation algorithm; adaptive multidimensional version; finite time performance; multidimensional newsvendor problem; scaled-and-shifted Kiefer-Wolfowitz algorithm; tuning parameter adjustment; Algorithm design and analysis; Analytical models; Approximation algorithms; Computational modeling; Convergence; Finite difference methods; Multidimensional systems; Performance analysis; Stochastic processes; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Simulation Conference (WSC), Proceedings of the 2009 Winter

Conference_Location

Austin, TX

Print_ISBN

978-1-4244-5770-0

Type

conf

DOI

10.1109/WSC.2009.5429708

Filename

5429708