Kennard-Stone method outperforms the Random Sampling in the selection of calibration samples in SNPs and NIR data
ABSTRACT: Splitting the whole dataset into training and testing subsets is a crucial part of optimizing models. This study evaluated the influence of the choice of the training subset in the construction of predictive models, as well as on their validation. For this purpose we assessed the Kennard-Stone (KS) and the Random Sampling (RS) methods in near-infrared spectroscopy data (NIR) and marker data SNPs (Single Nucleotide Polymorphisms). It is worth noting that in SNPs data, there is no knowledge of reports in the literature regarding the use of the KS method. For the construction and validation of the models, the partial least squares (PLS) estimation method and the Bayesian Lasso (BLASSO) proved to be more efficient for NIR data and for marker data SNPs, respectively. The evaluation of the predictive capacity of the models obtained after the data partition occurred through the correlation between the predicted and the observed values, and the corresponding square root of the mean squared error of prediction. For both datasets, results indicated that the results from KS and RS methods differ statistically from each other by the F test (P-value < 0.01). The KS method showed to be more efficient than RS in practically all repetitions. Also, KS method has the advantage of being easy and fast to be applied and also to select the same samples, which provides excellent benefits in the following analyses.
Main Authors: | , , |
---|---|
Format: | Digital revista |
Language: | English |
Published: |
Universidade Federal de Santa Maria
2022
|
Online Access: | http://old.scielo.br/scielo.php?script=sci_arttext&pid=S0103-84782022000500202 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
id |
oai:scielo:S0103-84782022000500202 |
---|---|
record_format |
ojs |
spelling |
oai:scielo:S0103-847820220005002022021-10-27Kennard-Stone method outperforms the Random Sampling in the selection of calibration samples in SNPs and NIR dataFerreira,Roberta de AmorimTeixeira,GabrielyPeternelli,Luiz Alexandre data division PLS regression BLASSO chemometrics prediction power ABSTRACT: Splitting the whole dataset into training and testing subsets is a crucial part of optimizing models. This study evaluated the influence of the choice of the training subset in the construction of predictive models, as well as on their validation. For this purpose we assessed the Kennard-Stone (KS) and the Random Sampling (RS) methods in near-infrared spectroscopy data (NIR) and marker data SNPs (Single Nucleotide Polymorphisms). It is worth noting that in SNPs data, there is no knowledge of reports in the literature regarding the use of the KS method. For the construction and validation of the models, the partial least squares (PLS) estimation method and the Bayesian Lasso (BLASSO) proved to be more efficient for NIR data and for marker data SNPs, respectively. The evaluation of the predictive capacity of the models obtained after the data partition occurred through the correlation between the predicted and the observed values, and the corresponding square root of the mean squared error of prediction. For both datasets, results indicated that the results from KS and RS methods differ statistically from each other by the F test (P-value < 0.01). The KS method showed to be more efficient than RS in practically all repetitions. Also, KS method has the advantage of being easy and fast to be applied and also to select the same samples, which provides excellent benefits in the following analyses.info:eu-repo/semantics/openAccessUniversidade Federal de Santa MariaCiência Rural v.52 n.5 20222022-01-01info:eu-repo/semantics/articletext/htmlhttp://old.scielo.br/scielo.php?script=sci_arttext&pid=S0103-84782022000500202en10.1590/0103-8478cr20201072 |
institution |
SCIELO |
collection |
OJS |
country |
Brasil |
countrycode |
BR |
component |
Revista |
access |
En linea |
databasecode |
rev-scielo-br |
tag |
revista |
region |
America del Sur |
libraryname |
SciELO |
language |
English |
format |
Digital |
author |
Ferreira,Roberta de Amorim Teixeira,Gabriely Peternelli,Luiz Alexandre |
spellingShingle |
Ferreira,Roberta de Amorim Teixeira,Gabriely Peternelli,Luiz Alexandre Kennard-Stone method outperforms the Random Sampling in the selection of calibration samples in SNPs and NIR data |
author_facet |
Ferreira,Roberta de Amorim Teixeira,Gabriely Peternelli,Luiz Alexandre |
author_sort |
Ferreira,Roberta de Amorim |
title |
Kennard-Stone method outperforms the Random Sampling in the selection of calibration samples in SNPs and NIR data |
title_short |
Kennard-Stone method outperforms the Random Sampling in the selection of calibration samples in SNPs and NIR data |
title_full |
Kennard-Stone method outperforms the Random Sampling in the selection of calibration samples in SNPs and NIR data |
title_fullStr |
Kennard-Stone method outperforms the Random Sampling in the selection of calibration samples in SNPs and NIR data |
title_full_unstemmed |
Kennard-Stone method outperforms the Random Sampling in the selection of calibration samples in SNPs and NIR data |
title_sort |
kennard-stone method outperforms the random sampling in the selection of calibration samples in snps and nir data |
description |
ABSTRACT: Splitting the whole dataset into training and testing subsets is a crucial part of optimizing models. This study evaluated the influence of the choice of the training subset in the construction of predictive models, as well as on their validation. For this purpose we assessed the Kennard-Stone (KS) and the Random Sampling (RS) methods in near-infrared spectroscopy data (NIR) and marker data SNPs (Single Nucleotide Polymorphisms). It is worth noting that in SNPs data, there is no knowledge of reports in the literature regarding the use of the KS method. For the construction and validation of the models, the partial least squares (PLS) estimation method and the Bayesian Lasso (BLASSO) proved to be more efficient for NIR data and for marker data SNPs, respectively. The evaluation of the predictive capacity of the models obtained after the data partition occurred through the correlation between the predicted and the observed values, and the corresponding square root of the mean squared error of prediction. For both datasets, results indicated that the results from KS and RS methods differ statistically from each other by the F test (P-value < 0.01). The KS method showed to be more efficient than RS in practically all repetitions. Also, KS method has the advantage of being easy and fast to be applied and also to select the same samples, which provides excellent benefits in the following analyses. |
publisher |
Universidade Federal de Santa Maria |
publishDate |
2022 |
url |
http://old.scielo.br/scielo.php?script=sci_arttext&pid=S0103-84782022000500202 |
work_keys_str_mv |
AT ferreirarobertadeamorim kennardstonemethodoutperformstherandomsamplingintheselectionofcalibrationsamplesinsnpsandnirdata AT teixeiragabriely kennardstonemethodoutperformstherandomsamplingintheselectionofcalibrationsamplesinsnpsandnirdata AT peternelliluizalexandre kennardstonemethodoutperformstherandomsamplingintheselectionofcalibrationsamplesinsnpsandnirdata |
_version_ |
1756406619365703680 |