STOCHASTIC GPU-BASED MULTITHREAD IMPLEMENTATION OF MULTIPLE BACK-PROPAGATION

Noel Lopes, Bernardete Ribeiro

2010

Abstract

Graphics Processing Units (GPUs) have evolved into a highly parallel, multi-threaded, many-core processor with enormous computational power. The GPU is especially well suited to address pattern recognition problems that can be expressed as data-parallel computations. Thus it provides a viable alternative to the use of dedicated hardware in the neural network (NN) field, where the long training times have always been a major drawback. In this paper, we propose a GPU implementation of the online (stochastic) training mode of the Multiple Back-Propagation (MBP) algorithm and compare it with corresponding standalone CPU version and with the batch training mode GPU implementation. For a fair and unbiased comparison we run the experiments with benchmarks from machine learning and pattern recognition field and we show that the GPU performance excel the CPU results in particular for high complex problems.

Paper Citation

in Harvard Style

Lopes N. and Ribeiro B. (2010). STOCHASTIC GPU-BASED MULTITHREAD IMPLEMENTATION OF MULTIPLE BACK-PROPAGATION . In Proceedings of the 2nd International Conference on Agents and Artificial Intelligence - Volume 1: ICAART, ISBN 978-989-674-021-4, pages 271-276. DOI: 10.5220/0002722102710276

in Bibtex Style

@conference{icaart10,
author={Noel Lopes and Bernardete Ribeiro},
title={STOCHASTIC GPU-BASED MULTITHREAD IMPLEMENTATION OF MULTIPLE BACK-PROPAGATION},
booktitle={Proceedings of the 2nd International Conference on Agents and Artificial Intelligence - Volume 1: ICAART,},
year={2010},
pages={271-276},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0002722102710276},
isbn={978-989-674-021-4},
}

in EndNote Style

TY - CONF
JO - Proceedings of the 2nd International Conference on Agents and Artificial Intelligence - Volume 1: ICAART,
TI - STOCHASTIC GPU-BASED MULTITHREAD IMPLEMENTATION OF MULTIPLE BACK-PROPAGATION
SN - 978-989-674-021-4
AU - Lopes N.
AU - Ribeiro B.
PY - 2010
SP - 271
EP - 276
DO - 10.5220/0002722102710276