Developing insight into effective SPSA parameters through sensitivity analysis

Ioulia Markou, Constantinos Antoniou

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Scopus citations

Abstract

Traffic simulation models have seen increasing use during the past decades. One of the biggest challenges related to their successful application, is the appropriate set of their values, thus achieving the accurate representation of driving and travel behavior parameters' diversity. Models' calibration using optimization algorithms, and more specifically the Simultaneous Perturbation Stochastic Approximation (SPSA) algorithm, is a crucial step. In this research, we study several aspects of SPSA's algorithm. A sensitivity analysis is implemented, in the context of finding the appropriate set of parameters, that will significantly improve its performance. Through successive experiments, the most efficient set is selected, and some guidelines are presented.

Original languageEnglish
Title of host publication2015 International Conference on Models and Technologies for Intelligent Transportation Systems, MT-ITS 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages58-65
Number of pages8
ISBN (Electronic)9789633131428
DOIs
StatePublished - 25 Aug 2015
Externally publishedYes
EventInternational Conference on Models and Technologies for Intelligent Transportation Systems, MT-ITS 2015 - Budapest, Hungary
Duration: 3 Jun 20155 Jun 2015

Publication series

Name2015 International Conference on Models and Technologies for Intelligent Transportation Systems, MT-ITS 2015

Conference

ConferenceInternational Conference on Models and Technologies for Intelligent Transportation Systems, MT-ITS 2015
Country/TerritoryHungary
CityBudapest
Period3/06/155/06/15

Keywords

  • SPSA
  • calibration
  • optimization
  • sensitivity analysis

Fingerprint

Dive into the research topics of 'Developing insight into effective SPSA parameters through sensitivity analysis'. Together they form a unique fingerprint.

Cite this