Open AccessMethod of paralleling grid nodes for multidimensional regions when solving parametric identification tasks
Author(s) -
Olga Kantor,
С. И. Спивак,
G. N. Yusupova,
S. L. Podvalny
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1479/1/012091
Subject(s) - grid , computer science , parametric statistics , process (computing) , identification (biology) , context (archaeology) , multidimensional data , multidimensional analysis , decomposition , algorithm , mathematical optimization , data mining , mathematics , statistics , geometry , paleontology , ecology , botany , biology , operating system
One of the approaches to solving parametric identification problems is based on numerical methods. In fact, their use is reduced to the process of iterating over the elements of multidimensional areas of the desired parameters. Increasing the density of the introduced grid increases the amount of calculations and time spent on the process of their implementation. Increasing the density of the introduced grid increases the amount of calculations and time spent on the process of their implementation. In this regard, it is desirable to improve the efficiency of the computer systems used. An important direction in the context of the above is the development of procedures for parallelization of programs. The article presents an algorithm for searching the nodal points of multidimensional areas based on the principle of data decomposition. In regard to solving the problem of parametric identification of the diffuse Bass model of the spread of innovation, its practical implementation has been carried out.