Parallelization of Adaboost Algorithm on Intel MIC Architecture

DOI

10.2991/icaita-16.2016.61How to use a DOI?

Keywords

Intel MIC; adaboost; parallel computation

Abstract

The Adaboost algorithm plays important role in many machine learning applications. But the computation cost is real expensive when the candidate features are in large amount. In this paper, we introduce a parallel strategy of the Adaboost algorithm on Intel CPU+MIC system, where Intel MIC works as a coprocessor. Open MP directive was used to parallelize the program in both CPU and MIC. The paper achieved a speedup of 5.2 on CPU+MIC with respect to CPU alone.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Haibiao Luo
AU  - Haojie Yuan
AU  - Shendong Cheng
AU  - Ying Li
AU  - Feng Yuan
AU  - Mingzhu Wei
PY  - 2016/01
DA  - 2016/01
TI  - Parallelization of Adaboost Algorithm on Intel MIC Architecture
BT  - Proceedings of the 2016 International Conference on Artificial Intelligence: Technologies and Applications
PB  - Atlantis Press
SP  - 248
EP  - 250
SN  - 1951-6851
UR  - https://doi.org/10.2991/icaita-16.2016.61
DO  - 10.2991/icaita-16.2016.61
ID  - Luo2016/01
ER  -