Banca de DEFESA: RAYANE SANTOS LEITE

Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.
STUDENT : RAYANE SANTOS LEITE
DATE: 30/08/2024
TIME: 15:00
LOCAL: Remoto
TITLE:

Application of GAMLSS and Machine Learning techniques in the estimation and prediction of the volume of Eucalyptus spp.

KEY WORDS:

Asymmetry; Caatinga; forest experimentation; monoculture; Araripe Gypsum Hub-PE; wood volume.

PAGES: 98
BIG AREA: Ciências Agrárias
AREA: Agronomia
SUMMARY:

This study investigates the application of advanced statistical models and Machine Learning algorithms in predicting the volume of Eucalyptus spp. in the Araripe Gypsum Hub region, Pernambuco. Generalized Additive Models for Location, Scale, and Shape (GAMLSS) were employed to model the volume of three Eucalyptus spp. clones across five different spacings, using a completely randomized design. The SHASHo model was selected as the most suitable due to its ability to handle asymmetric and dispersed data. Subsequently, a two-stage modeling approach was implemented to identify the most productive clone and spacing, utilizing the SHASH model and analysis of variance. Finally, Machine Learning algorithms, with a focus on XGBoost, were applied to predict the volume of the clones. XGBoost excelled with the best predictive performance, with Shapley value analysis identifying DBH and total height as the most influential variables. The integration of these methodologies proved effective in predicting volume and supporting the sustainable management of Eucalyptus spp. plantations, offering a robust alternative to meet the region's energy demands and reduce pressure on the Caatinga.

COMMITTEE MEMBERS:
Externo à Instituição - EDCARLOS MIRANDA DE SOUZA - UFAC
Externa à Instituição - ANA PATRÍCIA BASTOS PEIXOTO - UEPB
Externa ao Programa - 3303230 - CRISTIANE ROCHA ALBUQUERQUE - nullPresidente - JOSE ANTONIO ALEIXO DA SILVA
Externo à Instituição - MARCELINO ALVES ROSA DE PASCOA - UFMT