Banca de QUALIFICAÇÃO: RAUL NEGRÃO DE LIMA

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
STUDENT : RAUL NEGRÃO DE LIMA
DATE: 28/02/2022
TIME: 15:00
LOCAL: Remota
TITLE:

Basal area projection with artificial neural networks in managed gaps

KEY WORDS:

 artificial intelligence - sustainability - forest management - cutting cycle

PAGES: 50
BIG AREA: Ciências Agrárias
AREA: Recursos Florestais e Engenharia Florestal
SUMMARY:

Sustainable forest management is a great alternative for the Amazon against illegal logging. After the wood is harvested, clearings are created in the forest and studies show that just abandoning these areas is not the most correct option, aiming at the viability of the next cutting cycles. . Currently, research proves that the performance of post-harvest silvicultural treatments are potentiating the development of species with greater commercial value in the clearings. Thus, it is essential to use technologies that help predict the development of individuals in managed clearings in the Amazon. In this sense, the objective of the present project is to evaluate the projection of basal area in managed clearings, using artificial neural networks. For this research, data from the sustainable forest management area of Vale do Jari will be used. The neural networks will be tested using different configurations and as input the diameters collected in the measurements carried out in 2007 and 2008 will be used, with the basal area in 2009 as the output. in 2010 and 2011 data, targeting the basal area in 2012. The applied networks will be evaluated through the dispersion of percentage errors, square root of the mean error and application of the F test of Graybill.

BANKING MEMBERS:
Presidente - 741.673.590-49 - GUSTAVO SCHWARTZ - EMBRAPA
Interno - 142.060.489-91 - JOSE NATALINO MACEDO SILVA - Oxford
Externo à Instituição - DEIVISON VENICIO SOUZA
Externo à Instituição - LUIZ FERNANDES SILVA DIONISIO - UFRA