Publisher:
Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio), Departamento de Economia, Rio de Janeiro
Abstract:
The main goal of this paper is to apply a combination of statistical and connectionist schemes to examine, via clustering analysis, private indebtedness in different countries. Thirty-nine such experiences are used. The relationship between private debts and some macroeconomic variables are discussed in some detail. The clustering performance is improved by taking advantage of specific properties and capacities of each method. The procedures are also applied to a controlled numerical example.