Mixed zeros inflated poisson model: an application in grape disease control

Authors

  • Rafaela Pereira
  • Lucas Garcia Pedroso
  • Luiz Carlos Matioli

DOI:

https://doi.org/10.55905/oelv22n2-071

Keywords:

zero inflated poisson, mixed model, grape disease

Abstract

Longitudinal data studies describe the behavior of a feature of observations over time through so-called mixed models, which bear this name because they contain fixed effects and random effects. Still, if many of the observations do not show such a characteristic, i.e., the amount of zeros observed is greater than expected, then we say that it is inflated data of zeros. This work aims to describe the applicability of a mixed zero inflated Poisson (ZIP) model to model the temporal progress of mildew (Plasmopara viticultura), which corresponds to fungal disease of the grape, more specifically in the cultivar BRS Vitória, and assess its tolerance for 6 different treatments. This is intended to evaluate the quality of the proposed model fit and to carry out tests comparing the treatment averages to point out which treatments are equally effective in the control of mildew, thus demonstrating the application of the mixed ZIP model in a real and significant problem. A mixed ZIP model was constructed based on methodologies found in the literature and for inference and estimation of the parameters the method of maximum likelihood was used. The results showed that the mixed ZIP model adjusted well to the data, and hypothesis testing for average pointed to equality in efficacy between some treatment groups, thus making it possible to obtain a treatment with fewer sprays within each group.

References

BARBOSA, M. A. G.; FREITAS, D. M. S.; RIBEIRO JÚNIOR, P. M.; BATISTA, D. C. Doenças da videira. Informe Agropecuário, Lavras, v. 37, p. 86-98, 2016.

DAGNE, G. A. Bayesian semiparametric zero-inflated poisson model for longitudinal count data. Mathematical biosciences, Elsevier, v. 224, n. 2, p. 126–130, 2010.

FITZMAURICE, G. M.; LAIRD, N. M.; WARE, J. H. Applied Longitudinal Analysis. John Wiley & Sons, 2012.

HALL, D. B. Zero-inflated poisson and binomial regression with random effects: a case study. Biometrics, Wiley Online Library, v. 56, n. 4, p. 1030–1039, 2000.

KASSAHUN, W. et al. A zero-inflated overdispersed hierarchical poisson model. Statistical Modelling, SAGE Publications, v. 14, n. 5, p. 439–456, 2014.

LAMBERT, D. Zero-inflated poisson regression, with an application to defects in

manufacturing. Technometrics, Taylor & Francis, v. 34, n. 1, p. 1–14, 1992.

LIU, J.; MA, Y.; JOHNSTONE, J. A goodness-of-fit test for zero-inflated poisson mixed effects models in tree abundance studies. Computational Statistics & Data Analysis, Elsevier, v. 144, p. 106887, 2020.

PEREIRA, C. B. et al. Temporal dynamics and management of downy mildew on the table grape ‘brs vitória’in northern paraná. Semina: Ciências Agrárias, v. 39, n. 1, p.19–28, 2018.

RIDOUT, M.; DEMÉTRIO, C. G.; HINDE, J. Models for count data with many zeros. In:Proceedings of the XIXth international biometric conference. [S.l.: s.n.], 1998. v. 19, p.179–192.

SAS INSTITUTE. The NLMIXED Procedure. Disponível em: https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.5/statug/statug_nlmixed_overview01.htm. Acesso em: 05 jan. 2024.

WANG, K.; YAU, K. KW; LEE, Andy H. A zero-inflated Poisson mixed model to analyze diagnosis related groups with majority of same-day hospital stays. Computer methods and programs in biomedicine, v. 68, n. 3, p. 195-203, 2002

Published

2024-02-14

How to Cite

Pereira, R., Pedroso, L. G., & Matioli, L. C. (2024). Mixed zeros inflated poisson model: an application in grape disease control. OBSERVATÓRIO DE LA ECONOMÍA LATINOAMERICANA, 22(2), e3186. https://doi.org/10.55905/oelv22n2-071

Issue

Section

Articles