Partial Volume Prediction Through Nonlinear Mixed Modeling

AUTOR(ES)

Nicoletti, Marcos Felipe

FONTE

Floresta Ambient.

DATA DE PUBLICAÇÃO

08/08/2019

RESUMO

ABSTRACT The objective of this study was to assess the prediction of partial volumes with nonlinear mixed modeling for Pinus taeda. The volume of 558 trees was measured. The four-parameter logistic model was used in its modified form for the nonlinear mixed approach and, for comparison, the 5th degree polynomial was used. In the mixed modeling, the random effects diameter, age and place were inserted. The statistical criteria used to assess the quality of the adjustment were the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), standard error of the estimate (Syx) and residual graphical analysis. Among the random effects analyzed, age obtained the best adjustment. However, to predict partial volumes, it was noticed that, regardless of the analyzed portion of the trunk, the 5th degree polynomial had the best estimates, with a mean standard error of 20.1% of the estimate compared to 51.8% of the logistic.

Documentos Relacionados

• Modeling the stylized facts in finance through simple nonlinear adaptive systems
• Modeling of stem form and volume through machine learning
• MULTILEVEL NONLINEAR MIXED-EFFECTS MODEL AND MACHINE LEARNING FOR PREDICTING THE VOLUME OF Eucalyptus SPP. TREES
• Genome association study through nonlinear mixed models revealed new candidate genes for pig growth curves
• Characterization of intestinal absorption of C-glycoside flavonoid vicenin-2 from Lychnophora ericoides leafs in rats by nonlinear mixed effects modeling