arrow arrow Russian journal of management, 2015, 3, . 3 (15)

<<     >>

Upon decision-making in alternative design problems


Dimitri Golenko-Ginzburg

Ph.D., Professor of the Department of Industrial Engineering and Management, Ben-Gurion University of the Negev (Beer-Sheva, Israel); e-mail: e-mail -. , Java-Script ;

Sergey Ljubkin

Ph.D., Professor of the Department of Project and Innovation Management, Moscow State University of Economics, Statistics and Informatics (Moscow, Russian Federation); e-mail: e-mail -. , Java-Script ;

Nitzan Swid

lecturer of Department of Industrial Engineering and Management, Ariel University (Ariel, Israel), lecturer of Department of Management, Faculty of Social Sciences, Bar-Ilan University (Ramat-Gan, Israel); e-mail: e-mail -. , Java-Script

Manuscript received: 30.04.2015. Revised: 05.05.2015. Accepted: 14.05.2015. Published online: 30.06.2015.

Abstract. One of the main problems in alternative network planning boils down to determining the optimal variant to carry out the considered simulated program. In this paper we will formulate the optimal variants choice criteria for the case of homogenous alternative networks which have been described in our publications [13].

Key words: homogenous alternative stochastic network, full and joint variants, optimal decisionmaking variant, multi-variant optimization, optimality indicator.

1. Introduction

While examining homogenous alternative networks the problem focuses on determining the full variant of a design program which is optimal from the viewpoint of a certain accepted criterion. The difference between stochastic and deterministic alternative models reveals itself in future utilization of the results of such multi-variant optimization. In deterministic alternative networks the optimal variant has to be executed regardless of any future conditions and circumstances; furthermore, it may be recommended to be adopted as a kind of master plan whilst controlling the process of a complicated system design. For stochastic networks, when each of the competing variants has a non-zero implementation probability, control problems become more complicated, since we are facing the additional indeterminacy as to the ways of reaching the ultimate programs targets. Taking into account information regarding the stochastic variants quality, which has been acquired by means of the

<<     >>