@InProceedings{BarbosaNomaKortFons:2015:EsExCl,
author = "Barbosa, David Pereira and Noma, Alexandre and Korting, Thales
Sehn and Fonseca, Leila Maria Garcia",
affiliation = "{} and {} and {Instituto Nacional de Pesquisas Espaciais (INPE)}
and {Instituto Nacional de Pesquisas Espaciais (INPE)}",
title = "Um Estudo Experimental com Classificadores baseados em
Regi{\~o}es e Perfis EVI",
booktitle = "Anais...",
year = "2015",
editor = "Gherardi, Douglas Francisco Marcolino and Arag{\~a}o, Luiz
Eduardo Oliveira e Cruz de",
pages = "880--887",
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 = "INPE is responsible for several projects, including PRODES and
TerraClass. Basically, PRODES provides annual maps corresponding
to annual deforestation in Amazonia Based on a deforestation map,
TerraClass provides a classification map for the deforested areas:
agriculture, pasture, forest, hydrography, urban, etc. In order to
build a classification map, manual classification is a cumbersome
and tedious work. In this sense, automatic or semi-automatic
approaches are highly desirable for classification of the
deforested areas. Previous work compared different approaches by
using TerraClass data from 2008 for binary classification:
agriculture and non-agriculture.The present paper extends the
previous work in three aspects: (1) by including recent TerraClass
data from 2010; (2) by treating the multiclass case by considering
3 or more classes; and (3) by considering an additional boosting
approach for classification. Specifically, SVM, OPF, Naive Bayes,
Decision Tree, Nearest Neighbors and a boosting technique are
compared by following a k-fold cross validation.",
conference-location = "Jo{\~a}o Pessoa",
conference-year = "25-29 abr. 2015",
isbn = "978-85-17-0076-8",
label = "169",
language = "pt",
organisation = "Instituto Nacional de Pesquisas Espaciais (INPE)",
ibi = "8JMKD3MGP6W34M/3JM47GK",
url = "http://urlib.net/ibi/8JMKD3MGP6W34M/3JM47GK",
targetfile = "p0169.pdf",
type = "Classifica{\c{c}}{\~a}o e minera{\c{c}}{\~a}o de dados",
urlaccessdate = "28 abr. 2024"
}