Sparse matrix classification on imbalanced datasets using convolutional neural networks
Please use this identifier to cite or link to this item:
http://hdl.handle.net/10347/21079
Files in this item
Metadata of this item
Title: | Sparse matrix classification on imbalanced datasets using convolutional neural networks |
Author: | Pichel Campos, Juan Carlos Pateiro López, Beatriz |
Affiliation: | Universidade de Santiago de Compostela. Centro de Investigación en Tecnoloxías da Información Universidade de Santiago de Compostela. Departamento de Electrónica e Computación Universidade de Santiago de Compostela. Departamento de Estatística, Análise Matemática e Optimización |
Subject: | Sparse matrix | Classification | Imbalance | Deep learning | CNN | Performance | |
Date of Issue: | 2019 |
Publisher: | IEEE |
Citation: | Pichel, J. and Pateiro-Lopez, B., 2019. Sparse Matrix Classification on Imbalanced Datasets Using Convolutional Neural Networks. IEEE Access, 7, 82377-82389 |
Abstract: | This paper deals with the class imbalance problem in the context of the automatic selection of the best storage format for a sparse matrix with the aim of maximizing the performance of the sparse matrix vector multiplication (SpMV) on GPUs. Our classi cation method uses convolutional neural networks (CNNs) and proposes several solutions to mitigate the bias toward the majority classes when the data are not balanced. First, the CNNs are trained using images that represent the sparsity pattern of the matrices, whose pixels are colored according to different matrix features. In addition, we introduce a new network called SpNet, which achieves better results than a standard network as AlexNet in terms of prediction accuracy even having a more simple architecture. Finally, sampling techniques and cost-sensitive methods have been studied to give more emphasis on minority classes. The experiments conducted show that our classi ers are able to select the best performing format 92.8% of the time, obtaining 98.3% of the maximum attainable SpMV performance.Acomparison to other state-of-the-art classi cation methods is also provided, demonstrating the bene ts of our proposal |
Publisher version: | https://doi.org/10.1109/ACCESS.2019.2924060 |
URI: | http://hdl.handle.net/10347/21079 |
DOI: | 10.1109/ACCESS.2019.2924060 |
E-ISSN: | 2169-3536 |
Rights: | © 2019 by the authors. Licensee IEEE. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/) |
Collections
-
- Modestya-Artigos [65]
- CiTIUS-Artigos [110]
- EC-Artigos [96]
- EIO-Artigos [197]