Parallel and accurate k-means algorithm on CPU-GPU architectures for spectral clustering - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Concurrency and Computation: Practice and Experience Année : 2021

Parallel and accurate k-means algorithm on CPU-GPU architectures for spectral clustering

Résumé

Summary k-Means is a standard algorithm for clustering data. It constitutes generally the final step in a more complex chain of high-quality spectral clustering. However, this chain suffers from lack of scalability when addressing large datasets. This can be overcome by applying also the k-means algorithm as a preprocessing task to reduce the input data instances. We propose parallel optimization techniques for the k-means algorithm on CPU and GPU. Particularly we use a two-step summation method with package processing to handle the effect of rounding errors that may occur during the phase of updating cluster centroids. Our experiments on synthetic and real-world datasets containing millions of instances exhibit a speedup up to 7 for the k-means iteration time on GPU versus 20/40 CPU threads using AVX units, and achieve double-precision accuracy with single-precision computations.
Fichier non déposé

Dates et versions

hal-03468039 , version 1 (06-12-2021)

Identifiants

Citer

Guanlin He, Stephane Vialle, Marc Baboulin. Parallel and accurate k-means algorithm on CPU-GPU architectures for spectral clustering. Concurrency and Computation: Practice and Experience, 2021, pp.e6621. ⟨10.1002/cpe.6621⟩. ⟨hal-03468039⟩
37 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More