Ajustes de modelos não lineares a dados de crescimento com estrutura de erros independentes e autoregressivos de primeira ordem-aplicações

Carregando...
Imagem de Miniatura

Data

2013

Autores

Sallum Neto, Farid [UNESP]
Carvalho, Lídia Raquel de [UNESP]
Mischan, Martha Maria [UNESP]

Título da Revista

ISSN da Revista

Título de Volume

Editor

Resumo

The objective of the present work was to study the growth of cows and rats, comparing the models of Gompertz, Richards, logistic, monomolecular and von Bertalanffy. Adjustments were made with and without autocorrelation in the residuals. The comparison between the models was done through the measures of quality adjustment AIC, BIC, mean prediction error and Akaike weight, second (Motulsky and Christopoulos, 2004). For cows, the model that best fit was the monomolecular followed by von Bertalanffy and Gompertz, now for mice, von Bertalanffy, Gompertz and Richards proved best. Settings with autocorrelation in the residuals is shown always best.
Objetivou-se com o presente trabalho estudar o crescimento de vacas e de ratos, comparando-se os modelos de Gompertz, Richards, logístico, monomolecular e Von Bertalanffy. Foram feitos ajustes com estrutura de erros independentes e autorregressivos de primeira ordem. A comparação entre os modelos foi feita através dos avaliadores de qualidade de ajustamento AIC, BIC, erro de predição médio e peso de Akaike, segundo (Motulsky and Christopoulos, 2004). Para as vacas, o modelo que melhor se ajustou foi o monomolecular, seguido do Von Bertalanffy e Gompertz; já para os ratos, Von Bertalanffy, Richards e Gompertz se mostraram melhores. Os ajustes com autocorrelação nos resíduos se mostraram sempre melhores.

Descrição

Palavras-chave

Nonlinear models, Autocorrelation in the residuals, Growth data, Modelos não lineares, Autocorrelação nos resíduos, Dados de crescimento

Como citar

Revista Brasileira de Biometria, v. 31, n. 4, p. 631-644, 2013.