A Generative Adversarial Network Structure for Learning with Small Numerical Data Sets
In recent years, generative adversarial networks (GANs) have been proposed to generate simulated images, and some works of literature have applied GAN to the analysis of numerical data in many fields, such as the prediction of building energy consumption and the prediction and identification of live...
Guardado en:
Autores principales: | , , , |
---|---|
Formato: | article |
Lenguaje: | EN |
Publicado: |
MDPI AG
2021
|
Materias: | |
Acceso en línea: | https://doaj.org/article/fe9b46b4b0724a4391e380d7671035eb |
Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
id |
oai:doaj.org-article:fe9b46b4b0724a4391e380d7671035eb |
---|---|
record_format |
dspace |
spelling |
oai:doaj.org-article:fe9b46b4b0724a4391e380d7671035eb2021-11-25T16:38:43ZA Generative Adversarial Network Structure for Learning with Small Numerical Data Sets10.3390/app1122108232076-3417https://doaj.org/article/fe9b46b4b0724a4391e380d7671035eb2021-11-01T00:00:00Zhttps://www.mdpi.com/2076-3417/11/22/10823https://doaj.org/toc/2076-3417In recent years, generative adversarial networks (GANs) have been proposed to generate simulated images, and some works of literature have applied GAN to the analysis of numerical data in many fields, such as the prediction of building energy consumption and the prediction and identification of liver cancer stages. However, these studies are based on sufficient data volume. In the current era of globalization, the demand for rapid decision-making is increasing, but the data available in a short period of time is scarce. As a result, machine learning may not provide precise results. Obtaining more information from a small number of samples has become an important issue. Therefore, this study aimed to modify the generative adversarial network structure for learning with small numerical datasets, starting with the Wasserstein GAN (WGAN) as the GAN architecture, and using mega-trend-diffusion (MTD) to limit the bound of virtual samples that the GAN generates. The model verification of our proposed structure was conducted with two datasets in the UC Irvine Machine Learning Repository, and the performance was evaluated using three criteria: accuracy, standard deviation, and <i>p</i>-value. The experiment result shows that, using this improved GAN architecture (WGAN_MTD), small sample data can also be used to generate virtual samples that are similar to real samples through GAN.Der-Chiang LiSzu-Chou ChenYao-San LinKuan-Cheng HuangMDPI AGarticlesmall datasetsvirtual samplegenerative adversarial networkTechnologyTEngineering (General). Civil engineering (General)TA1-2040Biology (General)QH301-705.5PhysicsQC1-999ChemistryQD1-999ENApplied Sciences, Vol 11, Iss 10823, p 10823 (2021) |
institution |
DOAJ |
collection |
DOAJ |
language |
EN |
topic |
small datasets virtual sample generative adversarial network Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999 |
spellingShingle |
small datasets virtual sample generative adversarial network Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999 Der-Chiang Li Szu-Chou Chen Yao-San Lin Kuan-Cheng Huang A Generative Adversarial Network Structure for Learning with Small Numerical Data Sets |
description |
In recent years, generative adversarial networks (GANs) have been proposed to generate simulated images, and some works of literature have applied GAN to the analysis of numerical data in many fields, such as the prediction of building energy consumption and the prediction and identification of liver cancer stages. However, these studies are based on sufficient data volume. In the current era of globalization, the demand for rapid decision-making is increasing, but the data available in a short period of time is scarce. As a result, machine learning may not provide precise results. Obtaining more information from a small number of samples has become an important issue. Therefore, this study aimed to modify the generative adversarial network structure for learning with small numerical datasets, starting with the Wasserstein GAN (WGAN) as the GAN architecture, and using mega-trend-diffusion (MTD) to limit the bound of virtual samples that the GAN generates. The model verification of our proposed structure was conducted with two datasets in the UC Irvine Machine Learning Repository, and the performance was evaluated using three criteria: accuracy, standard deviation, and <i>p</i>-value. The experiment result shows that, using this improved GAN architecture (WGAN_MTD), small sample data can also be used to generate virtual samples that are similar to real samples through GAN. |
format |
article |
author |
Der-Chiang Li Szu-Chou Chen Yao-San Lin Kuan-Cheng Huang |
author_facet |
Der-Chiang Li Szu-Chou Chen Yao-San Lin Kuan-Cheng Huang |
author_sort |
Der-Chiang Li |
title |
A Generative Adversarial Network Structure for Learning with Small Numerical Data Sets |
title_short |
A Generative Adversarial Network Structure for Learning with Small Numerical Data Sets |
title_full |
A Generative Adversarial Network Structure for Learning with Small Numerical Data Sets |
title_fullStr |
A Generative Adversarial Network Structure for Learning with Small Numerical Data Sets |
title_full_unstemmed |
A Generative Adversarial Network Structure for Learning with Small Numerical Data Sets |
title_sort |
generative adversarial network structure for learning with small numerical data sets |
publisher |
MDPI AG |
publishDate |
2021 |
url |
https://doaj.org/article/fe9b46b4b0724a4391e380d7671035eb |
work_keys_str_mv |
AT derchiangli agenerativeadversarialnetworkstructureforlearningwithsmallnumericaldatasets AT szuchouchen agenerativeadversarialnetworkstructureforlearningwithsmallnumericaldatasets AT yaosanlin agenerativeadversarialnetworkstructureforlearningwithsmallnumericaldatasets AT kuanchenghuang agenerativeadversarialnetworkstructureforlearningwithsmallnumericaldatasets AT derchiangli generativeadversarialnetworkstructureforlearningwithsmallnumericaldatasets AT szuchouchen generativeadversarialnetworkstructureforlearningwithsmallnumericaldatasets AT yaosanlin generativeadversarialnetworkstructureforlearningwithsmallnumericaldatasets AT kuanchenghuang generativeadversarialnetworkstructureforlearningwithsmallnumericaldatasets |
_version_ |
1718413115999125504 |