The current COVID-19 pandemic is affecting different countries in different ways. The assortment of reporting techniques alongside other issues, such as underreporting and budgetary constraints, makes predicting the spread and lethality of the virus a challenging task. This work attempts to gain a better understanding of how COVID-19 will affect one of the least studied countries, namely Brazil. Currently, several Brazilian states are in a state of lock-down. However, there is political pressure for this type of measures to be lifted. This work considers the impact that such a termination would have on how the virus evolves locally. This was done by extending the SEIR model with an on / off strategy. Given the simplicity of SEIR we also attempted to gain more insight by developing a neural regressor. We chose to employ features that current clinical studies have pinpointed has having a connection to the lethality of COVID-19. We discuss how this data can be processed in order to obtain a robust assessment.

/pdf/flattening-the-curves-on-off-lock-down-strategies-for-covid-2w71jx6v8s.pdf

Flattening the curves: on-off lock-down strategies for COVID-19 with an application to Brazi

/pdf/flattening-the-curves-on-off-lock-down-strategies-for-covid-3lmiwhlqlv.pdf

Flattening the curves: on-off lock-down strategies for COVID-19 with an application to Brazil

Computer models have been an important tool to determine soil bulk density. This soil property is fundamental to estimate soil carbon reserves and consequently to understand the global carbon cycle. The estimation of soil bulk density is not a trivial task since it demands an intensive and often impractical work. The purpose of this paper is to evaluate the performance of a pedotransfer function against an Artificial Neural Networks to estimate soil bulk density for soils at Brazilian biomes. The first one consists of a linear model composed of a Least Square method. The latter employs a robust committee of multilayer perceptron networks and a model selection procedure based on k-fold cross-validation. The data are composed of 3404 soil layers distributed in different Brazilian regions and with different uses. The proposed non-linear regressor presents higher precision when compared to the linear model, and requires less information to do so. Additionally, the developed solution brings to light the assumed relationship between soil bulk density and some soil chemical properties.

Brazilian Soil Bulk Density Prediction Based on a Committee of Neural Regressors

Interrogation system for optical sensor using filter bank and artificial neural network

The Covid-19 pandemic has required great efforts from the entire scientific community. However, pandemic control has only been possible with social isolation and the use of facial masks. Despite this, cases of agglomerations and people who refuse to use the masks are shown daily, making it difficult to control the disease. This article presents an approach using Artificial Neural Networks (ANN) to identify people who do not use face masks in an image database. The results obtained demonstrate that although the parallelism used in the Python language with the Numpy library is punctual, it positively impacts the developed approach’s execution time. 1. Introdução O avanço da Covid-19 pelo mundo tem exigido esforços cada vez maiores dos cientistas no desenvolvimento de uma vacina. Até o momento as únicas medidas que se mostraram eficazes no controle da doença são baseadas no isolamento social, melhora nos hábitos de higiene e utilização das máscaras faciais. No entanto, inúmeras pessoas tem burlado os controles sanitários não utilizando as máscaras faciais, realizando aglomerações e etc. Verificar se essas medidas tem sido respeitadas por toda a população é uma tarefa complicada e exigiria um esforço muito grande das autoridades. Neste contexto, métodos de visão computacional podem ser utilizados para auxiliar na identificação automática de aglomerações ou mesmo identificar se um determinado indı́viduo está ou não utilizando a máscara dentro do metrô ou de um ônibus. O uso de tais métodos no contexto urbano não é novidade, inúmeras cidades já utilizam sistemas de monitoramento por câmeras [Ullah et al. 2019]. Dentre as técnicas de visão computacional, as Redes Neurais Artificiais (RNA) são uma das técnicas mais utilizadas, pois nos últimos anos elas têm obtido resultados consideráveis para problemas altamente complexos [Szegedy et al. 2016]. Assim, este trabalho propõe avaliar os impactos do paralelismo presente na biblioteca NumPy para implementação de uma RNA utilizada para identificar se uma determinada pessoa está utilizando a máscara facial. Este trabalho está estruturado conforme descrito a seguir. A Seção 2 apresenta os trabalhos relacionados. A Seção 3 apresenta uma pequena introdução sobre RNAMultiLayer Perceptron (MLP) e a metodologia desenvolvida. Os resultados obtidos são apresentados na Seção 4. Por fim, a Seção 5 apresenta as considerações finais. 2. Trabalhos Relacionados Com o avanço dos casos de Covid-19 pelo mundo, as contribuições cientı́ficas têm sido as mais abrangentes possı́veis. Neste contexto, diversas pesquisas demonstraram a efetividade do uso de máscaras faciais para a redução das taxas de transmissão do coronavı́rus ou até mesmo o grau de severidade da doença [Chughtai et al. 2020]. Contudo, as autoridades sanitárias têm tido cada vez mais dificuldade em fiscalizar o uso das máscaras por parte da população. Visando realizar tal fiscalização de forma automática, diversos pesquisadores têm proposto abordagens utilizando técnicas de visão computacional que identificam se um individuo está utilizando a máscara ou não[Ejaz et al. 2019, QIN and LI 2020, Loey et al. 2020]. A técnica de análise de componentes principais (PCA) foi utilizada por [Ejaz et al. 2019]. Já as Redes Neurais Artificiais foram utilizadas por [QIN and LI 2020] e [Loey et al. 2020]. Enquanto em [QIN and LI 2020] utilizaram redes do tipo Deep Super Resolution Crack Network (SRCNet), já as redes do tipo Residual Neural Network (ResNet50) foram avaliadas em [Loey et al. 2020]. Aqui são utilizadas as redes do tipo MLP, sendo avaliado o desempenho obtido com o uso do paralelismo da linguagem Python por meio da biblioteca NumPy.

/pdf/avaliacao-do-paralelismo-em-python-para-otimizar-uma-1xjc9qqpwq.pdf

Avaliação do Paralelismo em Python para Otimizar uma Abordagem de Identificação de Máscaras Faciais utilizando Redes Neurais Artificiais

Forest areas have an important role in the global carbon cycle. They are usually drains of greenhouse gases. From estimates of the soil bulk density Db, it is possible to determine the soil carbon stock. This relation shows the importance of determining D b . The purpose of this paper is to evaluate the performance of pedotransfer functions generated by a Linear Artificial Neural Network (ANN) to estimate D b for soils at Amazon Forest area. The raw data used was soil organic carbon (SOC), porosity, pH in water, sum of basic cations (SB) as predictor, Al3+ and hydrogen (H). These data were taken from a soil survey report conducted in the Oil Province of the Urucu River, composed by 378 samples. The model proposed present good results when compared with other approaches presented in related works.

A first approach using neural network to estimating soil bulk density of Urucu basin in Central Amazon-Brazil

The search for sustainable agricultural production system has required the monitoring of the quality of agricultural soils as well as the generation of modern tools for mapping the spatial variability of factors affecting the productivity of crops. The purpose of this work was to develop and evaluate the performance of an algorithm for mapping the spatial variability of soil physical quality, and develop a user-friendly and free software to perform the algorithm. Basically, the method consisted on: (1) Mapping the spatial variability of Easily Available Water (EAW), Air Filled Porosity (PA) and soil Penetration Resistance (PR), using geostatistics; and (2) Perform the integration of the maps of EAW, PA and PR into a map of Soil Physical Quality classes (SPQ), using Euclidean distance. The algorithm was tested in an agroecological farm with 6.24 ha, and was considered efficient to classify the SPQ. The attributes EAW and PA were the most important to define the SPQ of the study site. The software QUALIFIS met the demand for user-friendly software and can be applied in precision agriculture.

Rosane Ferreira de Oliveira

Papers

Brazilian Soil Bulk Density Prediction Based on a Committee of Neural Regressors

A first approach using neural network to estimating soil bulk density of Urucu basin in Central Amazon-Brazil

An Algorithm for Mapping the Spatial Variability of Soil Physical Quality