AKSELERASI K-MEANS CLUSTERING DENGAN COMPUTE UNIFIED DEVICE ARCHITECTURE (CUDA) (Studi Kasus pada SMA ISLAM HIDAYATULLAH)

GUSTA, WAHYU CEPTA (2014) AKSELERASI K-MEANS CLUSTERING DENGAN COMPUTE UNIFIED DEVICE ARCHITECTURE (CUDA) (Studi Kasus pada SMA ISLAM HIDAYATULLAH). Skripsi, Fakultas Ilmu Komputer.

Preview

PDF - Published Version Download (3604b) | Preview

Abstract

K-means algorithm is one of the unsupervised learning clustering algorithm that can be used to solve some problems such as marketing strategies, determining criteria schoolarship receiver, classifying documents until the identification vehicle number. A computation of sequential k-means algorithm is relatively high for at the time of execution. It is because of the computational processes executed according to a certain order before the next computation. To improve the performance of sequential k-means algorithm, parallelization of k-means algorithm is proposed with the aim to accelerate the response time of computational sequential k-means algorithm. On the other hand, the ability of a GPU can be used to do parallel computing. So that, the parallelization of k-means algorithm can be implemented by utilizing thousands of cores on the GPU.

Actions (login required)