Investigation of the Scheduler for Heterogeneous Distributed Computing Systems based on Minimal Cover Method

Listrovoy, S. V.; Minukhin, Sergii; Znakhur, S. V.

doi:10.5120/8154-1948

Cited by 1 publication

(1 citation statement)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It makes it possible to randomly generate the graph instances with a different number of vertices and a variable degree. The results of our method (the non-linear equations method) were compared with one of the most optimized methods (the frequency method) for the minimum vertex cover on the random graphs in heterogeneous distributed computing systems [16]. During the first group of tests the number of vertices of a given graph is gradually increased from 10 to 35 taking into account that the graph's density is fixed and has a value of 0.5.…”

Section: Resultsmentioning

confidence: 99%

The Probabilistic Method for Solving Minimum Vertex Cover Problem Using Systems of Non-linear Equations

Vladimirovich¹,

Vladimirovich²

2015

IJITCS

View full text Add to dashboard Cite

In this paper the probabilistic method is presented for solving the minimum vertex cover problem using systems of non-linear equations that are formed on the basis of a neighborhood relationship of a particular vertex of a given graph. The minimum vertex cover problem is one of the classic mathematical optimization problems that have been shown to be NP-hard. It has a lot of real-world applications in different fields of science and technology. This study finds solutions to this problem by means of the two basic procedures. In the first procedure three probabilistic pairs of variables according to the maximum vertex degree are formed and processed accordingly. The second procedure checks a given graph for the presence of the leaf vertices. Special software package to check the validity of these procedures was written. The experiment results show that our method has significantly better time complexity and much smaller frequency of the approximation errors in comparison with one of the most currently efficient algorithms.

show abstract

Section: Resultsmentioning

confidence: 99%