RESUMO Com a crescente produção de soja no Brasil e a demanda por grãos de soja com alto teor de proteína e óleo, é fundamental o estudo aprofundado dos constituintes desse grão, os quais podem variar de acordo com os genótipos e as condições de cultivo. Com isso, o objetivo desse estudo foi realizar a classificação de genótipos de soja, cultivados em diferentes ambientes e épocas de semeadura, de acordo com a composição química e o espectro gerado por infravermelho próximo (NIRs). Para isso, foi empregada a inteligência artificial e sua técnica de aprendizado de máquina. Foram utilizados 10 genótipos de soja, semeados em duas épocas de semeadura e em 7 cidades do Rio Grande do Sul. A composição química das amostras foi analisada através do equipamento FOSS NIRS DS2500, selecionando a banda entre 807 e 817nm. Os algoritmos aplicados foram J48, Random Forest, CVR, lBk, MLP, utilizando o filtro Resample. Foi empregado o software Weka, versão 3.8.6, para mineração de dados. O algoritmo IBk conseguiu o melhor desempenho, alcançando 89% de classificação correta dos atributos. A partir da Matriz de Confusão, observou-se que todos os genótipos obtiveram resultados superiores a 60/70 para os valores preditos corretamente, destacando o bom desempenho dos algoritmos. Nas métricas, o IBk obteve 0,89 de Precisão, Recall e F-Measure, e 0,94 de ROC Area. Foi possível classificar os genótipos, de acordo com a sua composição química relacionada aos dados obtidos na curva espectral, época e ambiente de semeadura, a partir da inteligência artificial e aprendizado de máquina. óleo grão cultivo isso NIRs. NIRs . (NIRs) máquina 1 Sul DS2500 DS 80 817nm nm J48 J Forest CVR lBk MLP Resample Weka 386 3 8 6 3.8.6 89 atributos Confusão observouse observou se 6070 60 70 60/7 corretamente métricas 089 0 0,8 Precisão FMeasure, FMeasure F Measure, Measure F-Measure 094 94 0,9 Area espectral (NIRs DS250 J4 38 3.8. 607 60/ 08 0, 09 9 DS25 3.8 DS2 3.

ABSTRACT With the increasing soybean production in Brazil, and the demand for soybeans with high protein and oil content, it is essential to conduct an in-depth study of the constituents of this grain, which can vary according to genotypes and growing conditions. Therefore, the objective of this study was to classify soybean genotypes, cultivated in different environments and sowing seasons, according to their chemical composition and the spectrum generated by near-infrared spectroscopy (NIRS). For this purpose, artificial intelligence and its machine learning technique were employed. 10 soybean genotypes were used, sown in two sowing seasons and cultivated 7 cities in Rio Grande do Sul. The chemical composition of the samples was analyzed using the FOSS NIRS DS2500 equipment, selecting the band between 807 and 817 nm. The applied algorithms were J48, Random Forest, CVR, lBk, MLP, using the Resample filter. The Weka software, version 3.8.6, was employed for data mining. The IBk algorithm achieved the best performance, reaching 89% correct classification of attributes. From the Confusion Matrix, it was observed that all genotypes obtained results above 60/70 for correctly predicted values, highlighting the algorithms’ good performance. In the metrics, IBk achieved 0.89 Precision, Recall, and F-Measure, and 0.94 ROC Area. Thus, it was possible to classify the genotypes according to their chemical composition related to the data obtained in the spectral curve, sowing season, and environment, using artificial intelligence and machine learning. Brazil content indepth depth grain conditions Therefore nearinfrared near infrared NIRS. . (NIRS) purpose 1 used Sul DS DS250 equipment 80 81 nm J48 J Forest CVR lBk MLP filter software 386 3 8 6 3.8.6 mining performance 89 attributes Matrix 6070 60 70 60/7 values metrics 089 0 0.8 Precision Recall FMeasure, FMeasure F Measure, Measure F-Measure 094 94 0.9 Area Thus curve season environment (NIRS DS25 J4 38 3.8. 607 60/ 08 0. 09 9 DS2 3.8 3.

Abstract To assess the physiological quality of soybean seeds, regulations require germination levels of at least 80% for a batch of seeds to be approved. With the evolution of technology and computing, spectrophotometric methods are advancing, making it possible to analyze the molecular vibrations present in seed samples and correlate them with germination results. The aim of this work was to analyze the use of Near Infrared (NIR) spectrometry with the aid of machine learning, in conjunction with physiological attributes, to predict the classification of soybean seed samples into accepted (above 80% germination) and rejected (below 79% germination), according to the legislation. A total of 134 soybean samples were analyzed for germination on paper, vigor by first count and tetrazolium. The NIR equipment was used to obtain the spectral curves of these samples. The Savitzky-Golay filter was used for pre-processing and data processing showed that the algorithm with the highest precision and accuracy was Random Forest, generating 96.1% recall and 98% average precision for the class of accepted lots, and 98.8% and 97.5% for the class of rejected lots. After using the algorithms, it was found that the range of 807.18 to 817.65nm can express information on physiological attributes and can be used to rank batches of soybean seeds, allowing for rapid estimation of the physiological quality of batches. Spectroscopy can be a complementary methodology with a quick response, with the potential to be used analogously to physiological quality analysis. 80 approved computing advancing results (NIR learning above below 79 germination, , legislation 13 paper tetrazolium SavitzkyGolay Savitzky Golay preprocessing pre Forest 961 96 1 96.1 98 lots 988 8 98.8 975 97 5 97.5 algorithms 80718 807 18 807.1 81765nm nm 817 65nm response analysis 7 9 96. 98. 97. 8071 807. 81

RESUMO A classificação de lotes de sementes é um processo fundamental para todas as empresas do setor sementeiro. O objetivo do trabalho é demonstrar os métodos de mineração de dados de ranqueamento de lotes de sementes de sorgo durante o processo de beneficiamento, através de análises de dados do controle de qualidade. Os testes realizados foram germinação e teste de frio, com o objetivo de verificar a qualidade fisiológica dos lotes. As amostras de sementes de cada lote foram avaliadas em dois momentos: póslimpeza e produto acabado (pronto para comercialização). Os dados gerados, após o pré-processamento, totalizaram 188 linhas com seis atributos, contabilizando 150 lotes aceitos para comercialização, seis rejeitados e 32 denominados intermediários. Os classificadores utilizados foram J48, Random Forest, Classification Via Regression, Naive Bayes, Multilayer Perceptron e IBk. Utilizou-se o filtro Resample para ajustamento dos dados. A técnica empregada para treinamento foi a k-fold, com 10 folds. Para verificar a precisão dos algoritmos foram utilizadas as métricas de Acurácia, Precisão, Recall, F-measure e Área ROC. Com os resultados obtidos determinou-se o melhor algoritmo de aprendizagem de máquina. Verificou-se que o IBk e o J48 obtiveram maior acurácia nos dados, sendo que a técnica de IBk obteve o melhor resultado. O filtro Resample foi importante para resolver o problema do desequilíbrio dos dados. Concluímos ser possível classificar lotes de sementes de sorgo com grande acurácia e precisão através de inteligência artificial e sua técnica de aprendizado de máquina. sementeiro beneficiamento frio momentos pronto comercialização. comercialização . comercialização) gerados préprocessamento, préprocessamento pré processamento, processamento pré-processamento 18 atributos 15 3 intermediários J Forest Regression Bayes Utilizouse Utilizou se kfold, kfold k fold, fold k-fold 1 folds Acurácia Precisão Recall Fmeasure F measure ROC determinouse determinou máquina Verificouse Verificou J4 resultado

ABSTRACT The ranking of seed lots is a fundamental process for all companies in the seed industry. This work aims to demonstrate data mining methods for ranking sorghum seed lots during the seed processing through analysis of quality control data. Germination and cold tests were performed to verify the physiological quality of the lots. Seed samples from each lot were evaluated in two moments: post-cleaning and finished product (ready for marketing). The results after pre-processing totaled 188 rows of data with six attributes, encompassing 150 lots accepted for marketing, 6 rejected, and 32 intermediate lots. The classifiers used were J48, Random Forest, Classification Via Regression, Naive Bayes, Multilayer Perceptron, and IBk. The Resample filter was used for adjustment of the data. The k-fold technique was used for training, with ten folds. The metrics of Accuracy, Precision, Recall, F-measure, and ROC Area were used to verify the accuracy of the algorithms. The results obtained were used to determine the best machine-learning algorithm. IBk and J48 presented the highest accuracy of data; the IBk technique presented the best results. The Resample filter was essential for solving the data imbalance problem. Sorghum seed lots can be classified with great accuracy and precision through artificial intelligence and machine learning technique. industry moments postcleaning post cleaning ready marketing. marketing . marketing) preprocessing pre 18 attributes 15 rejected 3 J Forest Regression Bayes Perceptron kfold k fold training folds Accuracy Precision Recall Fmeasure, Fmeasure F measure, measure F-measure algorithms machinelearning algorithm J4 problem 1

ABSTRACT To enhance speed and agility in interpreting physiological quality tests of seeds, The use of algorithms has emerged. This study aimed to identify suitable machine learning models to assist in the precise management of seed lot quality. Soybean lots from two companies were assessed using the Supplied Test Set, Cross-Validation (with 8, 10, and 12 folds), and Percentage Split (with 66% and 70%) methods. Variables analyzed through Tetrazolium tests included vigor, viability, mechanical damage, moisture damage, bed bug damage, and water content. Method performance was determined by Kappa, Precision, and ROC Area metrics. Classification Via Regression and J48 algorithms were employed. The technique utilizing 66% of data for training achieved 93.55% accuracy, with Precision and ROC Area reaching 94.50% for the J48 algorithm. Applying the cross-validation method with 10 folds resulted in 90.22% of correctly classified instances, with a ROC Area outcome like the previous method. Tetrazolium Vigor was the primary attribute used. However, these results are specific to this study's database, and careful planning is necessary to select the most effective application methods. seeds emerged Set CrossValidation Cross Validation 8 1 folds, , folds) 66 70% 70 methods vigor viability damage content Kappa metrics J J4 employed 9355 93 55 93.55 accuracy 9450 94 50 94.50 algorithm crossvalidation cross validation 9022 90 22 90.22 instances used However studys s database 6 7 935 9 5 93.5 945 94.5 902 2 90.2 93. 94. 90.

RESUMO Analisar os efeitos de dados de secagem em época de safra é primordial para o sucesso do armazenamento, além de manter a qualidade fisiológica das sementes. O objetivo com este trabalho foi avaliar o desempenho de secagem em secadores fixo e móveis através de aprendizado de máquinas. Os dados foram coletados de secadores convectivos, com base em: i) total de secadores trabalhados; ii) tempo de duração de secagem (horas); iii) percentual de umidade na entrada e saída do produto (%); e iv) diferenças de umidade entres ambas (%). Foi utilizado o modelo Filtered Clusterer, baseado na técnica Simple K-Means e filtro Resample, para agrupar dados com base em suas similaridades. Os resultados apresentados demonstram que o sistema de secagem é distinto entre os equipamentos fixo e móvel, com diferenças bem definidas dentro dos sistemas de secagem estabelecidos. O algoritmo juntamente com os filtros demonstrou ser eficiente na classificação não supervisionada, identificando e minimizando similaridade inter-cluster do sistema fixo definindo em classes distintas dentro do conjunto de dados. Conclui-se, que o conjunto de dados dispersos são agrupados de modo que o algoritmo classifica e minimiza similaridade inter-cluster do sistema fixo com alta precisão. Por outro lado, o desempenho da secagem em sistema móvel apresentou baixa eficiência. armazenamento sementes máquinas convectivos i trabalhados ii horas (horas) iii % (%) iv %. . Clusterer KMeans K Means Resample similaridades estabelecidos supervisionada intercluster inter cluster Concluise, Concluise Conclui se, se Conclui-se precisão lado eficiência (horas (% (

ABSTRACT Analyzing the impact of harvest-time drying data is crucial for successful storage and maintaining regulatory seed quality. This study aimed to assess the performance of fixed and mobile dryers using machine learning techniques. Data were collected from convective dryers, including the total number of dryers used, drying time (in hours), moisture percentages at the product’s entrance and exit, and the humidity difference between them. The study employed the Filtered Clusterer model, which utilizes the Simple K-Means technique and the Resample filter to group data based on similarities. The findings indicated distinct differences between fixed and mobile drying systems, with well-defined variations within each system. The algorithm, combined with the applied filters, proved effective in unsupervised classification by identifying and reducing inter-cluster similarity within the fixed system, thereby creating distinct classes within the dataset. In conclusion, the algorithm successfully clustered the scattered dataset and accurately classified and minimized inter-cluster similarity within the fixed system. Conversely, the mobile system exhibited low drying efficiency. harvesttime harvest quality techniques used hours, hours , hours) products product s exit them model KMeans K Means similarities systems welldefined well defined filters intercluster inter cluster conclusion Conversely efficiency

ABSTRACT The seed germination and vigor evaluation are essential for the sowing sector to measure the performance of different seed lots and improve the efficiency of storage and sowing processes. However, the analysis of various tests to determine seed quality generates a large amount of information, making it almost impossible for humans to perform a quick and effective quality control analysis. Therefore, the objective of this study was to evaluate the differences in the physiological quality of soybean seeds in different cultivars using machine learning techniques to rank the lots based on their quality. Three cultivars were used, and the analysis was germination, accelerated aging, tetrazolium treatment, seedling emergence, and 1000 seed weight from 65 lots were measured. The lots were evaluated in two phases, one immediately after harvest and the other after six months of storage. Random forest, multi-layer perceptron, J48, and classification via regression classifiers were used, aided by the feature resampler technique. Random forest and classification via regression obtained the highest accuracy, and the random forest technique obtained the best results. Therefore, it is possible to classify soybean seed lots with great accuracy and precision using artificial intelligence and machine learning techniques.