• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • Tagged with
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Mapeamento ótimo de doenças através da minimização simultânea do viés e da variância

Matos, Bárbara de Almeida e Silva Lima de 18 December 2012 (has links)
Dissertação (mestrado)—Universidade de Brasília, Instituto de Ciências Exatas, Departamento de Estatística, 2012. / Submitted by Alaíde Gonçalves dos Santos (alaide@unb.br) on 2013-03-15T14:10:13Z No. of bitstreams: 1 2012_BarbaradeAlmeidaeSilvaLimadeMatos.pdf: 1266319 bytes, checksum: b7ddb0567a162adb298f241b9bb1e409 (MD5) / Approved for entry into archive by Guimaraes Jacqueline(jacqueline.guimaraes@bce.unb.br) on 2013-03-20T11:47:04Z (GMT) No. of bitstreams: 1 2012_BarbaradeAlmeidaeSilvaLimadeMatos.pdf: 1266319 bytes, checksum: b7ddb0567a162adb298f241b9bb1e409 (MD5) / Made available in DSpace on 2013-03-20T11:47:04Z (GMT). No. of bitstreams: 1 2012_BarbaradeAlmeidaeSilvaLimadeMatos.pdf: 1266319 bytes, checksum: b7ddb0567a162adb298f241b9bb1e409 (MD5) / Propõe-se uma metodologia para gerar estimativas de taxa de doenças nas regiões de um mapa. A construção do estimador baseia-se na minimização simultânea de funções definidas como viés e variância local. O novo procedimento de estimação adota duas abordagens que consideram a informação espacial. As estimativas do novo procedimento são alcançadas por meio do algoritmo Multi-objective Particle Swarm Optimization (MOPSO). Foram realizadas comparações entre os resultados da nova metodologia e os provenientes do estimador bayesiano empírico local de Marshall. Conforme os critérios adotados neste trabalho, conclusões satisfatórias foram obtidas com o novo método e superaram os resultados do estimador bayesiano empírico local de Marshall. Uma aplicação com dados de câncer de mama na Nova Inglaterra, Nordeste dos Estados Unidos, é apresentada. _______________________________________________________________________________________ ABSTRACT / In this work it is proposed a methodology to produce estimates of disease rate in regions of a map. The construction of the estimator is based on the simultaneous minimization of functions defined as bias and local variance. The new estimation procedure adopts two approaches, both of which consider the spatial information. Estimates of the new procedure are achieved through the Multiobjective Particle Swarm Optimization (MOPSO) algorithm. The proposed methods were compared to Marshall’s local empirical Bayes estimator. According to the criteria adopted in this work, satisfactory conclusions were obtained and the results indicates that the new method is comparable to, or better than, Marshall’s local empirical Bayes estimator. An application to breast cancer data in New England, Northeast of the United States, is presented.

Page generated in 0.0765 seconds