Proposed Forest Prediction System based on Large-scale Adaptive Boosting Support Vector Regression Method
Abstract— In this paper, a forest prediction system for incorporating large-scale data on individual trees into one hybrid model is proposed. The proposed algorithm incorporates both forest biometry and statistical information, and constructs the hybrid model through combining adaptive boosting classification and support vector regression learning from large-scale forest data. More specifically, the species of a tree is firstly identified based on its measured features by using the adaptive boosting method. Subsequently, for each tree species the system relates the height of trees to the diameter at breast height and annual mean temperature for each tree species through a Support Vector Regression technique. This allows the tree’s height in the future to be well predicted. Experimental results show that the proposed algorithm has the capability to identify the species of trees and further predict tree growth through valid statistical inference.
Index Terms— forest prediction system, large-scale forest data, adaptive boosting method, Support Vector Machines, growth height of trees, statistical inference
Li-Li Wang, Matthew R Evans
The University of Hong Kong, Faculty of Science, HONG KONG
Cite: Li-Li Wang, Matthew R Evans, "Proposed Forest Prediction System based on Large-scale Adaptive Boosting Support Vector Regression Method," Proceedings of 2019 the 9th International Workshop on Computer Science and Engineering, pp. 143-149, Hong Kong, 15-17 June, 2019.