DOI

The paper is devoted to comparison of two methodologically different types of mathematical techniques for coping with network’s flows assignment problem. Gradient descent and projection approach are implemented to the simple network of parallel routes (there are no common arcs for any pair of routes). Gradient descent demonstrates zig-zagging behavior in some cases, while projection algorithm converge quadratically in the same conditions. Methodological interpretation of such phenomena is given.

Original languageEnglish
Title of host publicationLearning and Intelligent Optimization - 11th International Conference, LION 11, Revised Selected Papers
PublisherSpringer Nature
Pages345-350
Number of pages6
Volume10556 LNCS
ISBN (Print)9783319694030
DOIs
StatePublished - 2017
Event11th International Conference on Learning and Intelligent Optimization, LION 2017 - Nizhny Novgorod, Russian Federation
Duration: 18 Jun 2017 → 20 Jun 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10556 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference11th International Conference on Learning and Intelligent Optimization, LION 2017
Country/TerritoryRussian Federation
CityNizhny Novgorod
Period18/06/17 → 20/06/17

    Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

    Research areas

  • Gradient descent, Network’s flows assignment problem, Projection operator

ID: 10309639