@InProceedings{SouzaTelDruVolCun:2015:&bCaRe,
               author = "Souza, Vanessa Cristina Oliveira de and Tella, Bianca Gueldini and 
                         Drummond, Isabela Neves and Volpato, Margarete Marin Lordelo and 
                         Cunha, Rodrigo Luz da",
          affiliation = "{Instituto Nacional de Pesquisas Espaciais (INPE)}",
                title = "Aplica{\c{c}}{\~a}o de algoritmos de minera{\c{c}}{\~a}o de 
                         dados no reconhecimento de padr{\~o}es influentes na 
                         ocorr{\^e}ncia da ferrugem (Hemileia vastatrix berk. \&br) em 
                         cafeeiros na regi{\~a}o sul de Minas Gerais",
            booktitle = "Anais...",
                 year = "2015",
               editor = "Gherardi, Douglas Francisco Marcolino and Arag{\~a}o, Luiz 
                         Eduardo Oliveira e Cruz de",
                pages = "6874--6881",
         organization = "Simp{\'o}sio Brasileiro de Sensoriamento Remoto, 17. (SBSR)",
            publisher = "Instituto Nacional de Pesquisas Espaciais (INPE)",
              address = "S{\~a}o Jos{\'e} dos Campos",
             abstract = "Data mining techniques provide the recognition of patterns that 
                         determine or influence the disease infestation in coffee 
                         plantations.These models are very important to the strategic 
                         positioning of producers with respect to the rational use of 
                         pesticides and prevention measures. Frequent monitoring of large 
                         coffee crops is difficult and costly, since the manual data 
                         collection is restricted. As a proposed solution, this paper aims 
                         to develop a model of infestation of coffee rust (Hemileia 
                         vastatrix Berk. \& Br) directed to great land extension, using 
                         data from remote sensing (EVI) and data minig algorithms. We have 
                         employed decision trees and supervised neural network techniques 
                         to generate two models. One using meteorological variables and the 
                         other using EVI. The results corroborate the hypothesis that EVI 
                         can be replaced by meteorological variables in models of 
                         infestation in coffee rust in South of Minas Gerais. The models 
                         obtained accuracy of approximately 60%. The class ''high'' was the 
                         worst classified result obtained, due to the limited number of 
                         samples in the dataset.",
  conference-location = "Jo{\~a}o Pessoa",
      conference-year = "25-29 abr. 2015",
                 isbn = "978-85-17-0076-8",
                label = "1507",
             language = "pt",
         organisation = "Instituto Nacional de Pesquisas Espaciais (INPE)",
                  ibi = "8JMKD3MGP6W34M/3JM4JE6",
                  url = "http://urlib.net/ibi/8JMKD3MGP6W34M/3JM4JE6",
           targetfile = "p1507.pdf",
                 type = "Classifica{\c{c}}{\~a}o e minera{\c{c}}{\~a}o de dados",
        urlaccessdate = "27 abr. 2024"
}