Coffee Science
URI permanente desta comunidadehttps://thoth.dti.ufv.br/handle/123456789/3355
Navegar
3 resultados
Resultados da Pesquisa
Item Técnicas de extração de conhecimentos aplicadas à modelagem de ocorrência da cercosporiose (Cercospora coffeicola Berkeley & Cooke) em cafeeiros na região sul de Minas Gerais(Editora UFLA, 2013-01) Souza, Vanessa Cristina Oliveira de; Cunha, Rodrigo Luz da; Andrade, Lívia Naiara; Volpato, Margarete Marin Lordelo; Carvalho, Vicente Luiz de; Esmin, Ahmed Ali AbdallaO levantamento do progresso da cercosporiose torna-se potencialmente útil e compreensível no entendimento da doença e no processo de tomada de decisão para medidas de controle. Nos últimos anos, programas computacionais têm ajudado a elucidar quais fatores bióticos ou abióticos são mais representativos. Objetivou-se, neste trabalho, investigar, utilizando técnicas de extração do conhecimento, quais atributos ambientais e fenológicos mais influenciam na ocorrência da cercosporiose em cafeeiros no Sul de Minas Gerais, sob dois sistemas de cultivo: convencional e orgânico. Para isso, foram organizados dados de incidência de cercosporiose nos dois sistemas de cultivo, com dados climáticos e fenológicos da cultura, em um período de cinco anos de avaliação. Em seguida, um algoritmo de extração do conhecimento baseado em árvore de decisão foi utilizado para obter os atributos que mais favorecem a ocorrência da cercosporiose. Os modelos gerados tiveram 60% de taxa de acerto e mostraram que a temperatura média foi o atributo de maior influência na totalidade dos dados e para o sistema convencional de cultivo. No manejo orgânico, a precipitação mensal e a fenologia são os fatores que mais interferem na ocorrência da doença.Item Knowledge extraction techniques applied to blotch (Cercospora coffeicola Berkeley & Cooke) occurance modeling in coffee trees in southern Minas Gerais(Editora UFLA, 2013-01) Souza, Vanessa Cristina Oliveira de; Cunha, Rodrigo Luz da; Andrade, Lívia Naiara; Volpato, Margarete Marin Lordelo; Carvalho, Vicente Luiz de; Esmin, Ahmed Ali AbdallaBlotch progress lifting becomes potentially useful and understandable in understanding the disease and in the process of decision taking for control measures. In the last years, computer programs have been helping to elucidate which biotic or abiotic factors are more representative. The objective, in this study, was to investigate, using knowledge extraction technique, which environmental and phenological attributes influence the most in blotch occurrence in coffee trees in Southern Minas Gerais, under two crop systems: conventional and organic. To do so, blotch incidence data were organized on two crop systems, with climatic and phenological culture data, in five years of evaluation. Following that, a knowledge extraction algorithm based on decision tree was used to obtain the attributes which favor blotch occurrence the most. Generated models obtained 60% hit rate and showed that the average temperature was the most influential factor in total data and for conventional crop system. In organic management, monthly precipitation and phenology are the factors that interfere the most in disease occurrence.Item Geothecnologies in the assessment of land use changes in coffee regions of the state of Minas Gerais in Brazil(Editora UFLA, 2007-07) Vieira, Tatiana Grossi Chquiloff; Alves, Helena Maria Ramos; Bertoldo, Mathilde Aparecida; Souza, Vanessa Cristina Oliveira deCoffee is one of Brazilian most important cash crops due to the incomes generated by exportation. Southern Minas Gerais State represents approximately half of the total national production, although the greatest expansion of the crop has been observed in the western part of the state ( Triângulo Mineiro and Alto Paranaíba regions). As a requirement for future planning, it is important to establish efficient methodologies to map and monitor these lands, with the possibility of an easier periodical updating of the information. In this work geotechnologies were used to evaluate changes, in space and time, of areas occupied by coffee plantations in Minas Gerais. Land use maps of study areas selected in the main producing regions of the state were generated for the years 2000 and 2003 using the GIS SPRING and Landsat images. The results of the quantitative comparison of these maps indicated different behaviours for the regions evaluated. In São Sebastião do Paraíso and Machado, countries (south of Minas Gerais), a decrease of the areas occupied by coffee was observed, whereas in Três Pontas, increased. In Patrocínio, western part of the state, the area occupied by the crop remained unaltered. Remote sensing and GIS were efficient in the evaluation of the spatial-temporal dynamics of coffee lands of Minas Gerais, providing a greater understanding of the different environments and information that can support regional land use planning.