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.

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.

RESUMO OBJETIVO Avaliar a evolução da disponibilidade domiciliar de alimentos regionais no Amazonas, sua distribuição segundo características sociodemográficas e potenciais diferenças em relação ao restante do Brasil. MÉTODOS Foram analisados dados de aquisição de alimentos para consumo domiciliar das Pesquisas de Orçamentos Familiares (POFs) de 2002–2003, 2008–2009 e 2017–2018, que estudaram, respectivamente, 48.470, 55.970 e 57.920 domicílios no Brasil, dos quais 1,075, 1.344 e 1.833 estão no Amazonas. Os alimentos foram reunidos em três grupos: macaxeira e derivados, peixes de água doce, e frutos regionais. A quantidade de alimento regional adquirido, expressa em participação calórica relativa domiciliar, foi analisada para o Amazonas como um todo e segundo variáveis sociodemográficas (diferenças avaliadas pela sobreposição dos intervalos de confiança de 95%). RESULTADOS A participação calórica domiciliar do total de alimentos regionais no Amazonas foi de 22,54% em 2002-2003, 18,18% em 2008-2009 e 6,49% em 2017-2018. No Brasil, estes percentuais foram bem menores no mesmo período: 3,67%, 3,34% e 1,82%, respectivamente. As mudanças no Amazonas ocorreram, principalmente, pela drástica redução do grupo de macaxeira e derivados (de 14,30% em 2002–2003 para 12,74% em 2008-2009 e 3,09% em 2017–2018) e pelo declínio gradativo da disponibilidade domiciliar de peixes de água doce (de 7,30% em 2002–2003 para 4,85% 2008–2009 e 2,90% em 2017–2018). Domicílios do meio rural e com menor renda per capita tiveram maior participação calórica do total de alimentos regionais, estratos que também tiveram as maiores reduções. CONCLUSÃO Houve redução significativa da presença de alimentos regionais no Amazonas no período estudado, atingindo principalmente os domicílios da zona rural e com menor renda, cuja pessoa de referência da família era do sexo masculino, mais jovem e com menor escolaridade.

ABSTRACT OBJECTIVE To evaluate the evolution of household availability of regional foods in the state of Amazonas, their distribution according to sociodemographic characteristics, and potential differences when compared to the remaining areas of Brazil. METHODS Data on food acquisition for home consumption from the 2002-2003, 2008-2009, and 2017-2018 Pesquisa de Orçamentos Familiares (POFs – Consumer Expenditure Surveys) were analyzed, covering, respectively, 48,470, 55,970, and 57,920 households in Brazil, of which 1,075, 1,344, and 1,833 are in Amazonas. Foods were categorized into three groups: cassava and its derivatives, freshwater fish, and regional fruits. The study analyzed the amount of regional food purchased, expressed in relative household caloric share, for the entire area of Amazonas. Additionally, the data was stratified and analyzed according to sociodemographic variables, with differences assessed through the overlapping of 95% confidence intervals. FINDINGS The household caloric share of the total regional foods in Amazonas was 22.54% in 2002-2003, 18.18% in 2008-2009, and 6.49% in 2017-2018. Across Brazil, those percentages were much lower in the same period: 3.67%, 3.34%, and 1.82%, respectively. Changes in Amazonas were primarily attributed to the steep drop in the cassava and derivatives group, which decreased from 14.30% in 2002-2003 to 12.74% in 2008-2009 and further declined to 3.09% in 2017-2018. Additionally, there was a gradual decline in household availability of freshwater fish, decreasing from 7.30% in 2002-2003 to 4.85% in 2008-2009 and reaching 2.90% in 2017-2018. Households in rural areas and with lower per capita income presented a higher proportion of calories from total regional foods; this particular stratum also experienced the most significant reductions in their consumption. CONCLUSION During the study period, there was a significant decrease in the consumption of regional foods in Amazonas, particularly in lower income households in rural areas. Among them, the family reference person was typically a younger male with a lower educational background.

ABSTRACT The widespread adoption of sensor technology has made it a standard practice for obtaining precise and timely information during the harvest and post-harvest periods. One sensor that has gained popularity for post-harvest seed monitoring is the MQ-3, which identifies ethanol in the air as products undergo fermentation. However, these sensors typically require a stable operation. This study aimed to assess the stabilization time of an MQ-3 sensor when measuring ethanol levels in anaerobic bean seeds. We used six bean seed samples, each with an average moisture content of around 14%. We employed a completely randomized experimental design with nine repetitions for each sample. Every repetition consisted of 25 bean seeds placed in sealed flasks containing 70 mL of distilled water. This setup induced anoxic conditions within the flask, promoting anaerobic respiration in the seeds. After 24 hours, we exposed an air sample to the MQ-3 sensor and took readings at various time intervals (12-14, 19-21, 36-38, 68-70, 130-132, 192-194, 314-316, 616-618 seconds). The average stabilization time for the MQ-3 sensor while quantifying ethanol concentrations in the bean samples were approximately 23 seconds. The sensor demonstrated efficacy, convenience, and rapidity in assessing ethanol levels in anaerobic bean seeds.

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.

ABSTRACT OBJECTIVE To evaluate the evolution of household availability of regional foods in the state of Amazonas, their distribution according to sociodemographic characteristics, and potential differences when compared to the remaining areas of Brazil. METHODS Data on food acquisition for home consumption from the 2002-2003, 2008-2009, and 2017-2018 Pesquisa de Orçamentos Familiares (POFs – Consumer Expenditure Surveys) were analyzed, covering, respectively, 48,470, 55,970, and 57,920 households in Brazil, of which 1,075, 1,344, and 1,833 are in Amazonas. Foods were categorized into three groups: cassava and its derivatives, freshwater fish, and regional fruits. The study analyzed the amount of regional food purchased, expressed in relative household caloric share, for the entire area of Amazonas. Additionally, the data was stratified and analyzed according to sociodemographic variables, with differences assessed through the overlapping of 95% confidence intervals. FINDINGS The household caloric share of the total regional foods in Amazonas was 22.54% in 2002-2003, 18.18% in 2008-2009, and 6.49% in 2017-2018. Across Brazil, those percentages were much lower in the same period: 3.67%, 3.34%, and 1.82%, respectively. Changes in Amazonas were primarily attributed to the steep drop in the cassava and derivatives group, which decreased from 14.30% in 2002-2003 to 12.74% in 2008-2009 and further declined to 3.09% in 2017-2018. Additionally, there was a gradual decline in household availability of freshwater fish, decreasing from 7.30% in 2002-2003 to 4.85% in 2008-2009 and reaching 2.90% in 2017-2018. Households in rural areas and with lower per capita income presented a higher proportion of calories from total regional foods; this particular stratum also experienced the most significant reductions in their consumption. CONCLUSION During the study period, there was a significant decrease in the consumption of regional foods in Amazonas, particularly in lower income households in rural areas. Among them, the family reference person was typically a younger male with a lower educational background.

RESUMO OBJETIVO Avaliar a evolução da disponibilidade domiciliar de alimentos regionais no Amazonas, sua distribuição segundo características sociodemográficas e potenciais diferenças em relação ao restante do Brasil. MÉTODOS Foram analisados dados de aquisição de alimentos para consumo domiciliar das Pesquisas de Orçamentos Familiares (POFs) de 2002–2003, 2008–2009 e 2017–2018, que estudaram, respectivamente, 48.470, 55.970 e 57.920 domicílios no Brasil, dos quais 1,075, 1.344 e 1.833 estão no Amazonas. Os alimentos foram reunidos em três grupos: macaxeira e derivados, peixes de água doce, e frutos regionais. A quantidade de alimento regional adquirido, expressa em participação calórica relativa domiciliar, foi analisada para o Amazonas como um todo e segundo variáveis sociodemográficas (diferenças avaliadas pela sobreposição dos intervalos de confiança de 95%). RESULTADOS A participação calórica domiciliar do total de alimentos regionais no Amazonas foi de 22,54% em 2002-2003, 18,18% em 2008-2009 e 6,49% em 2017-2018. No Brasil, estes percentuais foram bem menores no mesmo período: 3,67%, 3,34% e 1,82%, respectivamente. As mudanças no Amazonas ocorreram, principalmente, pela drástica redução do grupo de macaxeira e derivados (de 14,30% em 2002–2003 para 12,74% em 2008-2009 e 3,09% em 2017–2018) e pelo declínio gradativo da disponibilidade domiciliar de peixes de água doce (de 7,30% em 2002–2003 para 4,85% 2008–2009 e 2,90% em 2017–2018). Domicílios do meio rural e com menor renda per capita tiveram maior participação calórica do total de alimentos regionais, estratos que também tiveram as maiores reduções. CONCLUSÃO Houve redução significativa da presença de alimentos regionais no Amazonas no período estudado, atingindo principalmente os domicílios da zona rural e com menor renda, cuja pessoa de referência da família era do sexo masculino, mais jovem e com menor escolaridade.

RESUMO No setor de produção de sementes, é imprescindível o uso de equipamentos rápidos e práticos que auxiliem nos processos de tomada de decisão. Assim, o uso de sensores que podem capturar informações para auxiliar na agricultura tem contribuído significativamente para o cultivo da soja em todo o mundo, sendo utilizado em diversas etapas, seja em pré ou pós-colheita. Objetivou-se desenvolver um protótipo de um etilômetro com sensor MQ-3 e avaliar seu tempo de estabilização quando utilizado em lotes de sementes de soja submetidos à hipóxia. A pesquisa foi realizada com 10 lotes de sementes de soja, oriundas de cinco cultivares (NS 8383 RR, SOY Peso RR, M 8372 IPRO, M8644 IPRO e M8808 IPRO), a qual foi dividida em duas etapas, a primeira correspondendo ao desenvolvimento do equipamento de aferição do etanol com o sensor MQ-3 e a segunda relacionada ao tempo de estabilização do sensor (3, 6, 9, 12, 15, 18, 21, 24, 27 e 30 s) em sementes de soja submetidas ao processo de anaerobiose. O equipamento, bem como o sensor MQ-3 apresentaram estabilidade de aferição média aos 16 s após seu acionamento, sendo recomendado esse tempo para o uso do etilômetro em sementes de soja. Por meio desse mesmo sensor MQ-3, é possível, com eficiência, a detecção de etanol presente no ar liberado por sementes de soja durante o processo de anaerobiose.

ABSTRACT In the seed production sector, it is essential to use fast and practical equipment to help decision-making processes. Thus, the use of sensors that can capture information to assist in agriculture has contributed significantly to soybean cultivation worldwide, being used in several stages, whether pre-harvest or post-harvest. Thus, this study aimed to develop a prototype of ethylometer with an MQ-3 sensor and to evaluate its stabilization time when used in soybean seed lots submitted to hypoxia. The research was carried out with 10 lots of soybean seeds from five cultivars (NS 8383 RR, SOY Peso RR, M 8372 IPRO, M8644 IPRO and M8808 IPRO) and was divided into two stages. The first one corresponds to the development of the ethanol measuring equipment with an MQ-3 sensor, and the second one to measurement of the sensor’s stabilization time (3, 6, 9, 12, 15, 18, 21, 24, 27, and 30 s) in soybean seeds submitted to anaerobiosis. The equipment and the MQ-3 sensor showed the stability of average measurement at 16 s after activation. This time is recommended for use of the ethylometer in soybean seeds. Through the MQ-3 sensor, it is possible to efficiently detect ethanol present in the air released by soy seeds during the anaerobic process.

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.

RESUMO: As sementes de coentro possuem alto valor socioeconômico em diversas regiões do Brasil, especialmente no Norte e Nordeste. A maturação de sementes determinada pela coloração influencia a qualidade da semente. Com isso, o processamento digital de imagens se tornou uma importante ferramenta para separação de sementes por meioda sua coloração, tendo em vista que essa separação normalmente é realizada por humanos e está altamente suscetível ao erro. O estudo tem como objetivo estabelecer parâmetros para separação de sementes de coentro por meio de processamento de análises de imagem red green and blue (RGB), buscando uma melhor seleção de sementes de coentro em função da sua coloração, assim como, avaliar a qualidade fisiológica por meio do teste de germinação. A partir de dois lotes de sementes de coentro, realizou-se a separação para obter amostras de 20g cada, em três tonalidades: amareladas, cinzas e mistas. As imagens foram adquiridas através da Scanner HP C4480 e processadas através do programa MATLAB, em seguida gerou-se um histograma para cada coloração analisada em cada amostra, por meio do sistema RGB. Na sequência, os resultados das médias das escalas foram analisados por ANOVA com a finalidade de ratificar a diferença das distribuições das componentes. Realizou-se o teste de germinação para ratificar os resultados da separação de sementes de coentro por meio de análises de imagens. A melhor seleção das sementes de coentro ocorreu através da escala em azul e o teste de germinação indicou que as sementes amarelas possuem qualidade fisiológica superior as sementes amarronzadas/acinzentadas.

ABSTRACT: Coriander seeds have high socio-economic value in several regions of Brazil, especially in the North and Northeast. Seed maturation determined by color influences the seed quality. With this, digital image processing has become an important tool for separating seeds by color since this classification is usually performed by humans and is highly susceptible to error. The study established parameters for separating coriander seeds by red green and blue (RGB) image analysis, seeking a better selection of coriander seeds according to their color, and evaluating the physiological quality by the germination test. Separation was carried out from two coriander seed lots to obtain samples of 20 g each in three shades: yellowish, gray, and mixed. Images were acquired by the HP C4480 Scanner and processed in the MATLAB software; then, a histogram was constructed for each color analyzed in each sample by the RGB system. ANOVA tested the averages of the scales to ratify the difference in the components’ distributions. The germination test was performed to confirm the results of seed separation using image analysis. The best selection of coriander seeds was achieved by the blue scale, and the germination test indicated that yellow seeds have a higher physiological quality than brownish/greyish seeds.

ABSTRACT The seed sector faces several challenges when it comes to ensuring a quick and accurate decision making when working with large amounts of data on physiological quality of seed lots, which makes the process time-consuming and inefficient. Thus, artificial intelligence (AI) emerges as a new technological option in the seed sector to solve database problems in the post-harvest stages. This study aims to use machine learning to classify maize seed lots. Data were obtained from eight maize seed crops from a private company. These data were mined using the following classifiers: J48 (DecisionTree), RandomForest, CVR (ClassificationViaRegression ) , lBk (lazy.IBK), MLP (MultiLayerPercepton), and NäiveBayes. Cross-validation was used for data measurement, with the data set, including training and testing data, being divided into 10 subsets. The described steps were performed using the Weka software. It is concluded that results obtained allow the classification of maize seed lots with high accuracy and precision, and these algorithms can better classify the maize seed lot through vigor attributes, thus enabling more accurate decision making based on vigor tests on a reduced evaluation time.

ABSTRACT New, efficient, low-cost techniques for image processing and alternative machine learning for seed processing are of academic and industrial interest. This study aims to identify fissures in bark and peeled rice seeds using X-ray and RGB image processing techniques and machine learning. Samples of three batches of rice seeds were used: a batch of seeds not subjected to drying (peeled seed), and the other two comprised of dried seeds, one containing seeds with husk and another containing huskless seeds; each sample comprised 100 seeds. Images in X-ray and RGB formats were provided in the sequence processed in ImageJ software and introduced in the machine learning software, where they were pre-processed using the appropriate filters and then classified by the J48 and linear discriminant analysis (LDA) classifiers. X-ray images obtained using differentiated equipment allow the identification of cracks in rice seeds using image processing techniques and the LDA classifier. Capturing images using RGB is a viable alternative. Using filters, either individually or in combination, may constitute an adequate alternative for rice seed classification.

RESUMO: No Brasil e no mundo, a comercialização da soja é de grande importância para a economia, sendo considerada sua qualidade. A presença de sementes danificadas de soja diminui o valor agregado do produto. As empresas precisam de técnicas rápidas e eficazes para manter a qualidade. O presente trabalho visa identificar, através do processamento de imagens, danos causados em sementes de soja, sendo eles: a presença de sementes esverdeadas e sementes enrugadas devido às variações de umidade e temperatura. Os resultados obtidos para sementes esverdeadas indicaram que a escala de cor vermelha é a mais adequada para seleção devido a sua variação mais significativa em relação às demais escalas de cores. Para a separação de sementes enrugadas, pode-se afirmar que é possível encontrar um parâmetro de seleção com 74,3% de acerto na remoção de sementes com médio a alto grau de danos por enrugamento.

ABSTRACT: In Brazil and worldwide, commercialization of soybeans is of great importance to the economy, making their quality considered. The presence of damaged soybean seeds decreases the added value of the product. Businesses need fast and effective techniques to maintain the quality. The present research aimed to identify, through image processing, damage caused to soybean seeds, namely the presence of greenish seeds and wrinkled seeds due to variations of humidity and temperature, where it was possible to identify greenish and wrinkled soybean seeds from images. Results obtained for greenish seeds indicated that the red color scale is the most suitable for selection due to its more significant variation compared to the other color scales. For the separation of wrinkled seeds, it can be stated that it is possible to find a selection parameter with 74.3% accuracy in removing seeds with medium to high degrees of wrinkle damage.

Este estudo teve como objetivo confirmar a estrutura e investigar as propriedades psicométricas de uma versão experimental portuguesa do York Cardiac Beliefs Questionnaire numa amostra da população geral; identificar as crenças erróneas mais fortes na comunidade; e avaliar as diferenças de acordo com características sociodemográficas. Trata-se de um estudo transversal com uma amostra de teste e outra de validação, incluindo um total de 476 participantes, com idade entre 18 e 40 anos, recrutados via e-mail e nas redes sociais. A Análise Fatorial Confirmatória em ambas as amostras indicou uma versão reduzida do YCBQ de três factores. As crenças erróneas diferiram significativamente de acordo com as variáveis sociodemográficas. A validação do YCBQ para amostras da comunidade constitui um importante ponto de partida para promover a investigação sobre crenças erróneas em grupos específicos da comunidade, assim como fornecer indicadores relevantes para a promoção da saúde.

This study sought to confirm the structure and to investigate the psychometric properties of an experimental Portuguese version of the York Cardiac Beliefs Questionnaire (YCBQ) in a general population sample. It also set out to identify the prevalent misconceptions in the community and to assess the differences according to socio-demographic characteristics. It involved a cross-sectional survey in which both test and validation samples were collected (n = 476), including participants aged between 18 and 40, recruited via e-mail and social networks. The Confirmatory Factor Analysis on both samples suggested a shorter, three factor version of the YCBQ. Also, misconceptions differed significantly according to sociodemographic variables. The validation of the YCBQ for samples in the community constitutes an important starting point to promote research on misconceptions held in the community by specific groups, as well as to provide key points for health promotion.

Este estudo teve como objetivo confirmar a estrutura e investigar as propriedades psicométricas de uma versão experimental portuguesa do York Cardiac Beliefs Questionnaire numa amostra da população geral; identificar as crenças erróneas mais fortes na comunidade; e avaliar as diferenças de acordo com características sociodemográficas. Trata-se de um estudo transversal com uma amostra de teste e outra de validação, incluindo um total de 476 participantes, com idade entre 18 e 40 anos, recrutados via e-mail e nas redes sociais. A Análise Fatorial Confirmatória em ambas as amostras indicou uma versão reduzida do YCBQ de três factores. As crenças erróneas diferiram significativamente de acordo com as variáveis sociodemográficas. A validação do YCBQ para amostras da comunidade constitui um importante ponto de partida para promover a investigação sobre crenças erróneas em grupos específicos da comunidade, assim como fornecer indicadores relevantes para a promoção da saúde.

This study sought to confirm the structure and to investigate the psychometric properties of an experimental Portuguese version of the York Cardiac Beliefs Questionnaire (YCBQ) in a general population sample. It also set out to identify the prevalent misconceptions in the community and to assess the differences according to socio-demographic characteristics. It involved a cross-sectional survey in which both test and validation samples were collected (n = 476), including participants aged between 18 and 40, recruited via e-mail and social networks. The Confirmatory Factor Analysis on both samples suggested a shorter, three factor version of the YCBQ. Also, misconceptions differed significantly according to sociodemographic variables. The validation of the YCBQ for samples in the community constitutes an important starting point to promote research on misconceptions held in the community by specific groups, as well as to provide key points for health promotion.