Comparação De Esquemas De Amostragem Para Treinamento De Modelos Preditores No Mapeamento Digital De Classes De Solos

Teske, Rodrigo; Giasson, Élvio; Bagatini, Tatiane

doi:10.1590/01000683rbcs20150344

Cited by 7 publications

(7 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Outros estudos com treinamento em mapas convencionais -como os de Giasson et al (2011) com acurácia global 68% e índice kappa 0,54), Teske et al (2015) com acurácia global 63% e índice kappa 0,46), e Silva et al (2013) com acurácia global 52 % e índice kappa 0,41 -também são comparáveis aos maiores valores de acurácia global (58%) e índice kappa (0,50) obtidos no presente estudo.…”

Section: Resultsunclassified

“…A capacidade preditiva dos modelos apoiados na mineração de dados também depende de bases de treinamento robustas, que captem ao máximo a variação dos atributos preditivos e dos solos (Teske et al, 2015). O treinamento por pontos de observação e coleta em campo cria elevada demanda por estes pontos para treinamento dos modelos, o que pode tornar o método pouco operacional.…”

Section: Introductionunclassified

See 1 more Smart Citation

Predição de classes de solo por mineração de dados em área da bacia sedimentar do São Francisco

Dias

Coelho

Valladares

et al. 2016

Pesq. agropec. bras.

View full text Add to dashboard Cite

Resumo -O objetivo deste trabalho foi avaliar diferentes estratégias para a predição da distribuição de classes de solo em mapas pedológicos digitais de áreas sem dados de referência, na bacia sedimentar do São Francisco, no Norte de Minas Gerais. As estratégias incluíram: o detalhamento da legenda, o treinamento por observações em campo, a ampliação do conjunto de treinamento e o uso de diferentes algoritmos de mineração de dados. Foram elaboradas quatro matrizes, diferenciadas pelo volume de dados, para o aprendizado dos algoritmos, e pelo nível taxonômico das classes de solo a serem preditas. Avaliou-se o desempenho dos algoritmos de aprendizado de máquina -Random Forest, J48 e MLP -, associados a procedimentos de discretização, balanceamento de classes, seleção de variáveis e expansão do conjunto de treinamento. O balanceamento de classes, a discretização de variáveis por frequências iguais e o algoritmo Random Forest apresentaram os melhores desempenhos. A extensão da representatividade das observações em campo, que presume uma área de treinamento mais ampla, não trouxe ganho preditivo. A generalização taxonômica para subordem diminui a fragmentação dos polígonos mapeados e aumenta a acurácia dos mapas pedológicos digitais. Quando são produzidos após treinamento por observações de solo in situ, na área de mapeamento, os mapas pedológicos digitais têm valores de acurácia equivalentes aos dos treinados em mapas preexistentes.Termos para indexação: acurácia de mapas pedológicos, algoritmos de classificação, mapa digital de solos, variáveis preditivas do meio físico. Soil class prediction by data mining in an area of the sedimentary São Francisco basinAbstract -The objective of this work was to evaluate different strategies for the prediction of soil class distribution on digital soil maps of areas without reference data, in the sedimentary basin of San Francisco, in the north of the state of Minas Gerais, Brazil. The strategies included: taxonomic generalization, training by field observations, training set expansion, and the use of different data mining algorithms. Four matrices were developed, differentiated by the volume of data for machine learning and by soil taxonomic levels to be predicted. The performance of the machine learning algorithms -Random Forest, J48, and MLP -, associated with discretization, class balancing, variable selection, and expansion of the training set was evaluated. Class balancing, variable discretization by equal frequencies, and the Random Forest algorithm showed the best performances. The representativeness extension of field observations, that assumes a larger training area, brought no predictive gain. Soil taxonomic generalization to the suborder level reduces the fragmentation of mapped polygons and improves the accuracy of digital soil maps. When generated by training on in situ soil observations at the mapping area, digital soil maps are as accurate as those trained on preexistent maps.

show abstract

Section: Resultsunclassified

Section: Introductionunclassified

Predição de classes de solo por mineração de dados em área da bacia sedimentar do São Francisco

Dias

Coelho

Valladares

et al. 2016

Pesq. agropec. bras.

View full text Add to dashboard Cite

show abstract

“…Of the 334 studies, 263 presented cartographic scale and spatial resolution (pixel size) information used (Table 3); 38 studies were found incompatible with the planimetric PEC-PCD, since their pixel size is higher than that indicated at the cartographic scale. (Giasson et al, 2006) 2008 (Figueiredo et al, 2008) 2009 (Crivelenti et al, 2009) 2010 (Chagas et al, 2010); (Coelho and Giasson, 2010) (Höfig et al, 2014); (Teske et al, 2014) 2015 (Bagatini et al, 2015); (Giasson et al, 2015); (Teske et al, 2015a); (Teske et al, 2015b); (Vasques et al, 2015) 2016 (Arruda et al, 2016); (Bagatini et al, 2016); (Demattê et al, 2016); (Dias et al, 2016); (Henrique et al, 2016); (Pelegrino et al, 2016) 2017 (Chagas et al, 2017); (Wolski et al, 2017) 2018 (Costa et al, 2018); (Meier et al, 2018) 2019 (Campos et al, 2019a); (Campos et al, 2019b); (Moura-Bueno et al, 2019); (Silva et al, 2019); (Silvero et al, 2019) All learning algorithms were assigned to a type of classifier such as Artificial Neural Network (ANN), Bayes classifiers, Decision Tree (DT), Logistic Regression (LR) and Support Vector Machine (SVM). Approximately 95 % of the studies used DT, ANN and LR classifiers (Table 4).…”

Section: Descriptive Statistics Of the Data Extracted From The Studiesmentioning

confidence: 99%

Digital soil class mapping in Brazil: a systematic review

Coelho

Giasson

Campos

et al. 2021

Sci. agric. (Piracicaba, Braz.)

Self Cite

View full text Add to dashboard Cite

In Brazil several digital soil class mapping studies were carried out from 2006 onwards to maximize the use of existing maps and information and to provide estimates for wider areas. However, there is no consensus on which methods have produced superior results in the predictive value of soil maps. This study conducts a systematic review of digital soil class mapping in Brazil and aims to analyze the factors which can improve the accuracy of digital soil class maps. Data from 334 digital soil class mapping studies were grouped and analyzed by Student's t-test, Wilcoxon-Mann-Whitney test and Kruskal-Wallis test. When conventional maps were used for validation, the studies showed average values of 63 % and when field samples were used, 56 % for Overall Accuracy. Studies compatible with the Planimetric Cartographic Accuracy Standard for Digital Cartographic Products (PEC-PCD) averaged between 4 % and 15 % higher accuracy than those of the incompatible group. There seems to be no evidence that increasing the number of variables and samples results in more accurate soil map prediction, but studies using variables related to four soil-forming factors enhanced accuracy. From a density of 0.08 MU km-2 and upwards, it became more difficult for studies to obtain greater accuracy. Artificial neural network classifiers and Decision Tree models seem to be producing more accurate digital soil class maps.

show abstract

“…Rev Bras Cienc Solo 2020;44:e0190120 Teske et al (2015), by using data from preexisting soil surveys, inferred that the sample design and the method of accuracy evaluation can affect the model predictor selection. Heung et al (2017) used a training dataset derived from soil pits and soil survey polygons to compare two types of machine learners, concluding that the RF model performed the best overall accuracy.…”

Section: Introductionmentioning

confidence: 99%

Sample design effects on soil unit prediction with machine: randomness, uncertainty, and majority map

Carvalho¹,

Pereira²,

Filho³

et al. 2020

Revista Brasileira De Ciência Do Solo

View full text Add to dashboard Cite

Notwithstanding the importance of soil surveys, advances in digital soil mapping have mainly focused on mapping soil attributes or properties rather than developing digital maps of soil units or soil classes. The purpose of this research was to develop digital soil unit maps based on primary soil data collection in areas without previously collected soil information. The covariate variability, the random effect across the data subset and the map outputs were the focuses of this study. We used five datasets with four models (Random Forest -RF, Gradient Boosted Machine -GBM, C5.0, and multinomial log-linear model -MLR). The covariates were grouped into five datasets, where four were grouped by Region Of Interest per Class (ROIC) and one was not grouped by ROIC. To evaluate the random effect to split the dataset, we ran each model 50 times and observed the overall accuracy (OA) and kappa index, and uncertainty, majority and variety maps. The OA of Dataset01 to 04 was lower than to Dataset05 accuracy. However, map outputs of RF and GBM for Dataset01 and Dataset05 had the same majority prediction. It seems that RF and GBM produce consistent results in map outputs according to this methodology and pedologist expertise. To evaluate the uncertainty and the consistency of soil unit prediction, we used the majority maps process. Random Forest, similar to GBM, presented the best results. The increase in the number of covariates was not a guarantee of improvement in the OA or in the quality of the map output. Geographic position and distance raster did not improve the map output according to expert evaluation. Because the variance between the ROICs, when the training and validation datasets were split based on it, the subsets are quite different in relation to the covariates, and this is the reason for the worse results of this model, comparing with the Dataset05. On the other hand, when considering one complete dataset not based on ROICs, the variance of training and validation subsets is lower and produced more accurate parameters of quality.

show abstract

Comparação De Esquemas De Amostragem Para Treinamento De Modelos Preditores No Mapeamento Digital De Classes De Solos

Cited by 7 publications

References 19 publications

Predição de classes de solo por mineração de dados em área da bacia sedimentar do São Francisco

Predição de classes de solo por mineração de dados em área da bacia sedimentar do São Francisco

Digital soil class mapping in Brazil: a systematic review

Sample design effects on soil unit prediction with machine: randomness, uncertainty, and majority map

Contact Info

Product

Resources

About