The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
A estratégia de implementação eficiente para acelerar algoritmos de agrupamento de alta qualidade é desenvolvida com base em unidades de processamento gráfico de uso geral (GPGPUs) neste trabalho. Entre vários algoritmos de agrupamento, um sofisticado modelo de mistura gaussiana (GMM), estimando parâmetros através do mecanismo bayesiano variacional (VB), é conduzido devido ao seu desempenho superior. Como a metodologia VB-GMM exige muita computação, o GPGPU é empregado para realizar cálculos matriciais massivos. Para migrar eficientemente os esquemas convencionais orientados a CPU do VB-GMM para plataformas GPGPU, um fluxo de migração completo com treze estágios é apresentado em detalhes. O esquema de cooperação CPU-GPGPU, reordenação de execução e otimização de acesso à memória são propostos para otimizar a utilização do GPGPU e maximizar a velocidade de clustering. Cinco tipos de aplicações do mundo real, juntamente com conjuntos de dados relevantes, são introduzidos para a validação cruzada. A partir dos resultados experimentais, verifica-se a viabilidade de implementação do algoritmo VB-GMM por GPGPU com benefícios práticos. A migração GPGPU proposta atinge uma aceleração máxima de 192x. Além disso, conseguiu identificar o número adequado de clusters, o que dificilmente é conduzido pelo algoritmo EM.
Hiroki NISHIMOTO
Nara Institute of Science and Technology
Renyuan ZHANG
Nara Institute of Science and Technology
Yasuhiko NAKASHIMA
Nara Institute of Science and Technology
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copiar
Hiroki NISHIMOTO, Renyuan ZHANG, Yasuhiko NAKASHIMA, "GPGPU Implementation of Variational Bayesian Gaussian Mixture Models" in IEICE TRANSACTIONS on Information,
vol. E105-D, no. 3, pp. 611-622, March 2022, doi: 10.1587/transinf.2021EDP7121.
Abstract: The efficient implementation strategy for speeding up high-quality clustering algorithms is developed on the basis of general purpose graphic processing units (GPGPUs) in this work. Among various clustering algorithms, a sophisticated Gaussian mixture model (GMM) by estimating parameters through variational Bayesian (VB) mechanism is conducted due to its superior performances. Since the VB-GMM methodology is computation-hungry, the GPGPU is employed to carry out massive matrix-computations. To efficiently migrate the conventional CPU-oriented schemes of VB-GMM onto GPGPU platforms, an entire migration-flow with thirteen stages is presented in detail. The CPU-GPGPU co-operation scheme, execution re-order, and memory access optimization are proposed for optimizing the GPGPU utilization and maximizing the clustering speed. Five types of real-world applications along with relevant data-sets are introduced for the cross-validation. From the experimental results, the feasibility of implementing VB-GMM algorithm by GPGPU is verified with practical benefits. The proposed GPGPU migration achieves 192x speedup in maximum. Furthermore, it succeeded in identifying the proper number of clusters, which is hardly conducted by the EM-algotihm.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2021EDP7121/_p
Copiar
@ARTICLE{e105-d_3_611,
author={Hiroki NISHIMOTO, Renyuan ZHANG, Yasuhiko NAKASHIMA, },
journal={IEICE TRANSACTIONS on Information},
title={GPGPU Implementation of Variational Bayesian Gaussian Mixture Models},
year={2022},
volume={E105-D},
number={3},
pages={611-622},
abstract={The efficient implementation strategy for speeding up high-quality clustering algorithms is developed on the basis of general purpose graphic processing units (GPGPUs) in this work. Among various clustering algorithms, a sophisticated Gaussian mixture model (GMM) by estimating parameters through variational Bayesian (VB) mechanism is conducted due to its superior performances. Since the VB-GMM methodology is computation-hungry, the GPGPU is employed to carry out massive matrix-computations. To efficiently migrate the conventional CPU-oriented schemes of VB-GMM onto GPGPU platforms, an entire migration-flow with thirteen stages is presented in detail. The CPU-GPGPU co-operation scheme, execution re-order, and memory access optimization are proposed for optimizing the GPGPU utilization and maximizing the clustering speed. Five types of real-world applications along with relevant data-sets are introduced for the cross-validation. From the experimental results, the feasibility of implementing VB-GMM algorithm by GPGPU is verified with practical benefits. The proposed GPGPU migration achieves 192x speedup in maximum. Furthermore, it succeeded in identifying the proper number of clusters, which is hardly conducted by the EM-algotihm.},
keywords={},
doi={10.1587/transinf.2021EDP7121},
ISSN={1745-1361},
month={March},}
Copiar
TY - JOUR
TI - GPGPU Implementation of Variational Bayesian Gaussian Mixture Models
T2 - IEICE TRANSACTIONS on Information
SP - 611
EP - 622
AU - Hiroki NISHIMOTO
AU - Renyuan ZHANG
AU - Yasuhiko NAKASHIMA
PY - 2022
DO - 10.1587/transinf.2021EDP7121
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E105-D
IS - 3
JA - IEICE TRANSACTIONS on Information
Y1 - March 2022
AB - The efficient implementation strategy for speeding up high-quality clustering algorithms is developed on the basis of general purpose graphic processing units (GPGPUs) in this work. Among various clustering algorithms, a sophisticated Gaussian mixture model (GMM) by estimating parameters through variational Bayesian (VB) mechanism is conducted due to its superior performances. Since the VB-GMM methodology is computation-hungry, the GPGPU is employed to carry out massive matrix-computations. To efficiently migrate the conventional CPU-oriented schemes of VB-GMM onto GPGPU platforms, an entire migration-flow with thirteen stages is presented in detail. The CPU-GPGPU co-operation scheme, execution re-order, and memory access optimization are proposed for optimizing the GPGPU utilization and maximizing the clustering speed. Five types of real-world applications along with relevant data-sets are introduced for the cross-validation. From the experimental results, the feasibility of implementing VB-GMM algorithm by GPGPU is verified with practical benefits. The proposed GPGPU migration achieves 192x speedup in maximum. Furthermore, it succeeded in identifying the proper number of clusters, which is hardly conducted by the EM-algotihm.
ER -