Academic Journal

Organization of branchings for solving unification problems by correcting algorithms

Bibliographic Details
Title: Organization of branchings for solving unification problems by correcting algorithms
Authors: Gol'dengorin, B. I.
Publisher Information: Scripta Technica, Inc., Silver Spring
Subject Terms: Combinatorial optimization, decomposition, Numerical mathematical programming methods, corrective algorithm scheme, unification, allocation, branch and bound, branching, Integer programming, discrete optimization, irregular initial data
Description: Some differences are explained concerning the corrective algorithm scheme published previously by the author [Sov. Math., Dokl. 27, 620-623 (1983; Zbl 0555.65042)] from the branch and bound classic scheme to solve the discrete optimization problems belonging to various classes. An algorithm is suggested with respect to branching (decomposition) into subproblems for the single class of unification and allocation type problems [see the author, ibid. 33, 578-583 (1986; Zbl 0636.65059)]. One type of strategy to unify problems for the specific classes of regular initial data consists of solving them during uptime with minimal absolute errors, and another strategy is directed to work out the unification problem with given absolute errors during the processing time. A possibility is shown to introduce a balanced pairs of errors received during the approximation of irregular initial data to eliminate them after carrying out the branching along corresponding pairs of rows. The article gives information about the possibility to apply the decomposition for realizing the different variants in branching in the scheme of branch and bound type.
Document Type: Article
File Description: application/xml
Access URL: https://zbmath.org/4087439
Accession Number: edsair.c2b0b933574d..17f6114a0d08a6f31fb332113909bda2
Database: OpenAIRE
Description
Description not available.