You are in the accessibility menu

Please use this identifier to cite or link to this item: http://acervodigital.unesp.br/handle/10400.2/2554
Title: 
Previsão eleitoral para a Assembleia da República Portuguesa
Author(s): 
ISSN: 
1646-6330
Abstract: 
  • Em trabalho anterior utilizaram-se técnicas de Data Mining para predizer resultados eleitorais, sem utilizar sondagens, recorrendo a variáveis socioeconómicas, disponíveis publicamente sobre Portugal, no período abrangido pelas treze eleições para a Assembleia da República, entre 1974 e 2009. No entanto, o espectro político considerado nesse trabalho não abrange os 100% dos votos expressos, mas apenas os quatro partidos com assento parlamentar regular desde 1975 cuja votação atinge cerca de 84%. Na abordagem anteriormente adoptada, cada um dos quatro partidos tradicionais foi tratado separadamente, resultando em previsões independentes. Neste artigo analisa-se a extensão desse trabalho à previsão do intervalo restante dos resultados eleitorais e a sua utilização para garantir a restrição de que a percentagem total de votos expressos soma 100%. Os resultados mostraram que os métodos anteriormente aplicados permitem obter previsões com resultados de qualidade similar para o conjunto das forças partidárias complementares.
  • In earlier work, Data Mining techniques using a large number of public available socioeconomic variables about Portugal were used to predict election results between 1974 and 2009 (thirteen elections) for the Parliament of the Portuguese Republic. However, the political spectrum considered in that work does not cover 100% of the votes cast, but only the four parties with regular parliamentary seat since 1975, whose voting reaches about 84%. In the approach previously adopted, each of the four traditional parties was treated separately resulting in independent forecasts. In this paper we analyse the extension of that work to the forecasting of the remaining interval of the election results and its use to ensure the restriction that the total percentage of votes cast adds up to 100%. The results showed that the methods previously applied allow making predictions with similar quality for the remaining partisan forces.
Issue Date: 
2012
Citation: 
Azevedo, Diamantino; Correia, Luís; Gaspar, Graça - Previsão eleitoral para a Assembleia da República Portuguesa. "Revista de Ciências da Computação" [Em linha]. ISSN 1646-6330. Vol. 7, nº 7 (2012), p. 43-56
Publisher: 
Universidade Aberta
Keywords: 
  • Correlação
  • Assembleia da República
  • Previsão eleitoral
  • Variáveis socioeconómicas
  • Correlation
  • Parliament
  • Electoral forecast
  • Socioeconomic variables
  • REA
URI: 
Rights: 
openAccess
Type: 
outro
Source:
http://repositorioaberto.uab.pt/handle/10400.2/2554
Appears in Collections:Recursos Educacionais Abertos (REA) - Universidade Aberta de Portugal

There are no files associated with this item.
 

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.