• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • Tagged with
  • 1
  • 1
  • 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

Regressão Logística Geograficamente Ponderada aplicada a modelos de Credit Scoring

Medina, Fabio Augusto Scalet 27 April 2016 (has links)
Dissertação (mestrado)—Universidade de Brasília, Faculdade de Economia, Administração e Contabilidade, Programa de Pós-Graduação em Administração, 2016. / Submitted by Fernanda Percia França (fernandafranca@bce.unb.br) on 2016-06-23T16:39:44Z No. of bitstreams: 1 2016_FabioAugustoScaletMedina.pdf: 980307 bytes, checksum: c57d891197aa6227064966be4883bc3e (MD5) / Approved for entry into archive by Raquel Viana(raquelviana@bce.unb.br) on 2016-06-24T21:35:55Z (GMT) No. of bitstreams: 1 2016_FabioAugustoScaletMedina.pdf: 980307 bytes, checksum: c57d891197aa6227064966be4883bc3e (MD5) / Made available in DSpace on 2016-06-24T21:35:55Z (GMT). No. of bitstreams: 1 2016_FabioAugustoScaletMedina.pdf: 980307 bytes, checksum: c57d891197aa6227064966be4883bc3e (MD5) / A presente dissertação de mestrado teve como objetivo principal verificar a aplicabilidade da metodologia Regressão Logística Geograficamente Ponderada (GWLR) para a construção de modelos de credit scoring. As fórmulas do melhor conjunto de modelos locais estimados via GWLR foram comparadas entre si, em termos de valor dos coeficientes e significância das variáveis, e frente ao modelo global estimado via Regressão Logística. Foram utilizados dados reais referentes às operações de Crédito Direto ao Consumidor (CDC) de uma instituição financeira pública nacional concedidas a clientes domiciliados no Distrito Federal (DF). Os resultados encontrados demonstraram a viabilidade da utilização da técnica GWLR para desenvolver modelos de credit scoring. Os modelos estimados para cada região do DF se mostraram distintos em suas variáveis e coeficientes (parâmetros) e três dos cinco indicadores do modelo via GWLR se mostraram superiores aos do modelo via Regressão Logística. ________________________________________________________________________________________________ ABSTRACT / This master thesis aimed to verify the applicability of the methodology Geographically Weighted Logistic Regression (GWLR) to develop credit scoring models. The formulas of the best set of local models estimated by GWLR were compared in terms of value of the coefficients and significance of the variables, and against the global model estimated by Logistic Regression. It was used a real granting data of Direct Credit Consumer from a national public financial institution to borrowers domiciled in the Federal District (FD) of Brazil. The results demonstrated the feasibility of using the technique GWLR to develop credit scoring models. The estimated models for each region of FD have showed to be different in their variables and coefficients (parameters) and three out of five indicators calculated for the developed model by GWLR were superiors than indicators of the developed model by Logistic Regression.

Page generated in 0.1135 seconds