Modelos de regressão na descrição do crescimento de frutos de amora-preta

• Ano de defesa: 2018
• Autor(a) principal: Silva, Édipo Menezes da
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Lavras; Programa de Pós-Graduação em Estatística e Experimentação Agropecuária; UFLA; brasil; Departamento de Ciências Exatas
• Palavras-chave em Português: Análise de regressão; Seleção de modelos; Comportamento sigmoide; Desenvolvimento de frutos; Regression analysis; Selection of models; Sigmoid behavior; Fruit development; Estatística
• Link de acesso: http://repositorio.ufla.br/jspui/handle/1/29692

The Blackberry is a small fruit with several properties beneficial to human health and its cultivation is an alternative for small producers due to the fact that it has good financial returns. Studying the fruit growth over time is extremely important to understand its development, helping to better manage the crop, avoiding, for example, post-harvest loss, which is one of the aggravating factors of blackberry losses, since it has a short period of development. Thus, growth curves are highlighted in this type of study and modeling through statistical models helps to understand how such growth happens. The data were obtained in an experiment carried out at the Federal University of Lavras in 2015. The linear model and the non-linear models Brody, Logistic, Gompertz, double logistic and double Gompertz models were adjusted with the inclusion of the first-order autoregressive term when necessary. The objective of this work was to adjust linear and nonlinear models to describe the diameter and length growth of four cultivars of blackberry (Brazos, Choctaw, Guarani and Tupi). The estimation of the parameters was obtained through the least squares methods using the Gauss-Newton method, in addition to the "nls" and "glns" functions of the statistical software R. The comparison of the adjustments was made by Akaike (AIC), Bayesian information criterion (BIC), residual standard deviation (DPR) and adjusted determination coefficient (R²aj). The models described satisfactorily the data, with predominance for the linear first-degree model and the logistic-diphasic model.