Please use this identifier to cite or link to this item:
metadata.artigo.dc.title: Procedure to identify outliers through cumulative distribution of extremes in a gamma response model
metadata.artigo.dc.creator: Resende, Mariana
Brighenti, Carla Regina Guimarães
Cirillo, Marcelo Ângelo
metadata.artigo.dc.subject: False negatives
False positives
Mahalanobis distance
metadata.artigo.dc.publisher: Taylor and Francis Online 2017
metadata.artigo.dc.identifier.citation: RESENDE, M.; BRIGHENTI, C. R. G.; CIRILLO, M. A. Procedure to identify outliers through cumulative distribution of extremes in a gamma response model. Communications in Statistics - Simulation and Computation, [S.l.], v. 46, n. 9, 2017.
metadata.artigo.dc.description.abstract: This work aimed at proposing a procedure based on the cumulative distribution of maximums and minimums to identify outliers in generalized Gamma-response models. In order to validate such method, we used simulations scenarios defined by the combination of different samples, contamination rate and distributions with different degrees of asymmetry. In this context, probabilities related to errors in classification and accuracy were obtained by carrying by Monte Carlo simulations. Using cumulative distribution of extremes to identify outliers in a Gamma-response model is recommended, since it is not likely to present errors and was highly accurate in all assessed scenarios.
metadata.artigo.dc.language: en_US
Appears in Collections:DES - Artigos publicados em periódicos

Files in This Item:
There are no files associated with this item.

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