Loading...
Publication
ParalelizaĆ§Ć£o do algoritmo K-means
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1.08 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Neste artigo sĆ£o exploradas as vantagens da paralelizaĆ§Ć£o do algoritmo K-means. O algoritmo Ć© implementado na linguagem de programaĆ§Ć£o C com a utilizaĆ§Ć£o de tarefas (threads) POSIX e sĆ£o analisadas as consequĆŖncias ao nĆvel do desempenho pela utilizaĆ§Ć£o de programaĆ§Ć£o multitarefa. O algoritmo K-means Ć© essencialmente iterativo e a paralelizaĆ§Ć£o deste permite tirar partido do mĆ”ximo de processadores disponĆveis e com isso reduzir consideravelmente o tempo de execuĆ§Ć£o. Este ganho no tempo total de execuĆ§Ć£o permite um investimento no aprimorar do algoritmo de forma a obter resultados mais precisos, ou processar um maior volume de dados e manter a execuĆ§Ć£o em tempo Ćŗtil.
This article explores the advantages of parallelizing the K-means algorithm. The algorithm will be implemented in the C programming language using POSIX Threads and the consequences of the use of multi-thread programming in terms of performance will be analysed. The K-means algorithm is essentially iterative, and its parallelization will make it possible to take advantage of as many processors as possible and thereby considerably reduce the execution time. This gain in the total execution time allows an investment in improving the algorithm to obtain more accurate results, or to process a larger volume of data and keep the execution time limited.
This article explores the advantages of parallelizing the K-means algorithm. The algorithm will be implemented in the C programming language using POSIX Threads and the consequences of the use of multi-thread programming in terms of performance will be analysed. The K-means algorithm is essentially iterative, and its parallelization will make it possible to take advantage of as many processors as possible and thereby considerably reduce the execution time. This gain in the total execution time allows an investment in improving the algorithm to obtain more accurate results, or to process a larger volume of data and keep the execution time limited.
Description
Keywords
k-means Pthreads Linguagem de programaĆ§Ć£o C ProgramaĆ§Ć£o paralela multitarefa C programming language Multitasking parallel programming
Pedagogical Context
Citation
Martins, Marco; Shirley, Paulo - ParalelizaĆ§Ć£o do algoritmo K-means. "Revista de CiĆŖncias da ComputaĆ§Ć£o" [Em linha]. ISSN 1646-6330 (Print) 2182-1801 (Online). Vol. 16 (2021), p. 81-92
Publisher
Universidade Aberta