A novel generalized odd log-logistic Maxwell-based regression with application to microbiology

Autor(es)

Prataviera, Fabio ^[1] ; Silva, Antonio M. M. ^[2] ; Cardoso, Elke J. B. N. ^[2] ; Cordeiro, Gauss M. ^[3] ; Ortega, Edwin M. M. ^[1]

Afiliação do(s) autor(es)

^[1] Univ Sao Paulo, Dept Exact Sci, Piracicaba, SP - Brazil

^[2] Univ Sao Paulo, Dept Soil Sci, Piracicaba, SP - Brazil

^[3] Univ Fed Pernambuco, Dept Stat, Recife, PE - Brazil

• Tipo de documento: Artigo Científico
• Fonte: Applied Mathematical Modelling; v. 93, p. 148-164, MAY 2021.
• Citações Web of Science: 0

Resumo

The presence of bimodality, heteroskedasticity, zero-inflation and nonlinear effects in co variables is common in several real data applications. In this context, new regressions are proposed for data with all these characteristics. The estimation of parameters follows the maximum likelihood method. For different fixed parameters, sample sizes and percentages of zeros, various simulations are performed to assess the behavior of the estimators. Quantile residuals are defined to evaluate the assumptions of the proposed regression. Its usefulness is illustrated by an experiment conducted to assess the soil microbiology in a sugarcane field. (c) 2020 Elsevier Inc. All rights reserved. (AU)

• Processo FAPESP: 16/18944-3 - Mudanças climáticas e eficiência energética na agricultura: um enfoque em estresse hídrico, manejo orgânico e biologia do solo
• Beneficiário: Elke Jurandy Bran Nogueira Cardoso
• Modalidade de apoio: Auxílio à Pesquisa - Programa de Pesquisa sobre Mudanças Climáticas Globais - Temático