We here propose a new implementation of the forward search, which is a powerful general method usually suitable to detect extreme observations and to determine their effect on fitted models (Atkinson and Riani, 2000). The forward search technique is applied in the field of the size distribution of business firms. In particular, a threshold is fixed for the fit of the Pareto II distribution through a progressive adaptation technique, performing at each iteration the chi-square test to check for the acceptance of the null hypothesis. Non-parametric confidence intervals are produced to assess the stability of the estimates.
Corbellini, A., Crosato, L., Ganugi, P. (2007). Robust sequential fitting of the Pareto II distribution: theoretical and computational aspects. In Classification and Data Analysis 2007. Book of short Papers. Meeting of the Classification and Data Analysis Group of the Italian Statistical Society. Macerata : Eum.
Robust sequential fitting of the Pareto II distribution: theoretical and computational aspects
CROSATO, LISA;
2007
Abstract
We here propose a new implementation of the forward search, which is a powerful general method usually suitable to detect extreme observations and to determine their effect on fitted models (Atkinson and Riani, 2000). The forward search technique is applied in the field of the size distribution of business firms. In particular, a threshold is fixed for the fit of the Pareto II distribution through a progressive adaptation technique, performing at each iteration the chi-square test to check for the acceptance of the null hypothesis. Non-parametric confidence intervals are produced to assess the stability of the estimates.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.