@InCollection{SouzaMKAVDSJB:2018:ClMaPa,
               author = "Souza, Vitor M. and Medeiros, Cl{\'a}udia and Koga, Daiki and 
                         Alves, Livia Ribeiro and Vieira, Lu{\'{\i}}s Eduardo Antunes and 
                         Dal Lago, Alisson and Silva, Ligia Alves da and Jauer, Paulo 
                         Ricardo and Baker, Daniel",
                title = "Classification of magnetospheric particle distributions via neural 
                         networks",
            booktitle = "Machine learning techniques for space weather",
            publisher = "Elsevier",
                 year = "2018",
               editor = "Camporeale, Enrico and Johnson, Jay and Wing, Simon",
                pages = "329--353",
             keywords = "Self-organizing map, Pitch angle distribution, Earth’s 
                         magnetosphere.",
             abstract = "In this chapter we introduce a special kind of neural network 
                         known as a self-organizing map (SOM) and use it to 
                         cluster/classify pitch angle-resolved particle flux data obtained 
                         by instruments onboard satellites orbiting the Earth. As an 
                         example of the technique, we employ electron flux data at both 
                         relativistic and subrelativistic energies provided by two 
                         instruments onboard one of the twin NASAs Van Allen Probes. For 
                         these data sets the SOM can identify the shapes of three 
                         well-known types of pitch angle distributions, and from that 
                         knowledge one can infer the associated physical mechanisms in the 
                         near-Earth space environment, particularly in the Van Allen 
                         radiation belts region. The SOM-based methodology can be used with 
                         multiplatform spacecraft data, thus enabling a prompt 
                         characterization of the physical processes throughout the Earths 
                         magnetosphere. The steps required to apply our neural 
                         network-based approach to pitch angle-resolved particle flux data 
                         from any spacecraft mission are laid out.",
          affiliation = "{Instituto Nacional de Pesquisas Espaciais (INPE)} and {Instituto 
                         Nacional de Pesquisas Espaciais (INPE)} and {Instituto Nacional de 
                         Pesquisas Espaciais (INPE)} and {Instituto Nacional de Pesquisas 
                         Espaciais (INPE)} and {Instituto Nacional de Pesquisas Espaciais 
                         (INPE)} and {Instituto Nacional de Pesquisas Espaciais (INPE)} and 
                         {Instituto Nacional de Pesquisas Espaciais (INPE)} and {Instituto 
                         Nacional de Pesquisas Espaciais (INPE)} and {University of 
                         Colorado Boulder}",
                  doi = "10.1016/B978-0-12-811788-0.00013-5",
                  url = "http://dx.doi.org/10.1016/B978-0-12-811788-0.00013-5",
                 isbn = "0128117893",
                label = "lattes: 2470949000200852 4 SouzaMKAVDSJB:2018:ClMaPa",
             language = "en",
           targetfile = "souza_classification.pdf",
        urlaccessdate = "27 abr. 2024"
}