Parallel Processing in Neural Network for Pattern Recognition

Lubis, Chairisni and Hiryanto, Lely and Sulianto, Kurniawan (2008) Parallel Processing in Neural Network for Pattern Recognition. Karya Ilmiah Dosen. ISSN 1823-3287

[img]
Preview
 Text
1365-3158-1-PB.pdf - Published Version
Download (235kB) | Preview

Abstract

Perceptron is one of the fundamental algorithms in neural network for pattern recognition. The common problem faced in recognizing pattern using neural network is in the learning computation which is time consuming. To reduce the learning time, we develop a parallel algorithm for the perceptron learning process. To support the parallel processing on the cluster, we use MPICH for windows as message passing tool and Visual C++ 6.0 as the editor and compiler of the parallel pattern recognition program. Our experimental result of running the program in the cluster shows that the parallel process gives speedup 3 times of its sequential process.

Item Type: Article
Subjects: Penelitian > Fakultas Teknologi Informasi
Divisions: Fakultas Teknologi Informasi > Teknik Informatika
Depositing User: Puskom untar untar
Date Deposited: 30 Mar 2017 03:14
Last Modified: 30 Mar 2017 03:14
URI: http://repository.untar.ac.id/id/eprint/201

Actions (login required)

View Item View Item
Repository UNTAR is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.