Estimation of height, volume and taper of eucalyptus trees using support vector machine
Name: LUANDSON ARAÚJO DE SOUZA
Publication date: 21/02/2017
Advisor:
Name |
Role |
|---|---|
| ADRIANO RIBEIRO DE MENDONÇA | Advisor * |
Examining board:
| Name |
Role |
|---|---|
| ADRIANO RIBEIRO DE MENDONÇA | Advisor * |
| GILSON FERNANDES DA SILVA | Internal Examiner * |
Summary: Faced with restrictions on the use of uneven-aged forests, the investment in even-aged forests has been increasing. This increases the interest in studies seeking to maximize the yield in volume of these forests. The growth of this sector has driven the search for more accurate methods in the estimation of dendrometric variables in order to obtain reliable estimates of the volumetric stock of the forests stands. This work had as objective to evaluate the accuracy of estimate of total height, volume and stem taper of eucalyptus trees, using Support Vector Machine (SVM). For the realization this work, were used data from a eucalyptus stand. The database were randomly divided into two groups: 70% for fit of regression models and training of MVS and 30% for validation. Were fitted different configurations of SVM and different regression models. The evaluation of the regression models and configurations of SVM was based on the statistics: correlation coefficient, root mean square error and bias. Based on the results obtained, it was observed that the SVM provided greater accuracy in the estimates of total height, individual volume and taper in relation to the classic regression models used.
Keywords: Regression Models, Machine Learning, Dendrometric Variables.
