Parallelization of data mining algorithms for multicore processors

Author

Kholod, Ivan ; Kuprianov, Mikhail ; Shorov, Andrey

Author_Institution

Fac. of Comput. Sci. & Technol., St. Petersburg Electrotech. Univ. "LETI", St. Petersburg, Russia

fYear

2015

fDate

14-18 June 2015

Firstpage

262

Lastpage

267

Abstract

The article describes a approach of parallel data mining algorithms to be executed on multicore processors of various architecture. The suggested method presents an algorithm as a consequence of pure functions with unified interfaces. For parallel execution additional functions are introduced to share data and models between the parallel threads. Besides such functions allow to obtain various parallel algorithm structures and implement various strategies of execution for different environment conditions. Application of the described method is illustrated through algorithm Naïve Bayes.

Keywords

Bayes methods; data mining; multiprocessing systems; parallel algorithms; multicore processors; naïve Bayes algorithm; parallel data mining algorithms; parallel threads; Algorithm design and analysis; Approximation algorithms; Data mining; Data models; Multicore processing; Parallel algorithms; Program processors; data mining; data mining algorithms; distributed data mining; parallel algorithms; parallel data mining;

fLanguage

English

Publisher

ieee

Conference_Titel

Embedded Computing (MECO), 2015 4th Mediterranean Conference on

Conference_Location

Budva

Print_ISBN

978-1-4799-8999-7

Type

conf

DOI

10.1109/MECO.2015.7181919

Filename

7181919