ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

Using Decision Tree and Random Forests to Classify Land Coverage in Tomine Reservoir

Journal: International Journal of Mechanical and Production Engineering Research and Development (IJMPERD ) (Vol.10, No. 5)

Publication Date:

Authors : ; ;

Page : 21-34

Keywords : : Thematic information extraction; Supervised classification; Decision Tree (DT); Random Forests (RF); Land Cover; Satellite Image; Thematic accuracy;

Source : Downloadexternal Find it from : Google Scholarexternal

Abstract

Traditional methods of cover classification should be reviewed and compared against the recent methods proposed for this purpose, so the purpose of this study was to extract thematic information from a Landsat 5 TM satellite image of Tominé Reservoir, and around using two methods of classification and / or regression: Decision Tree (DT) and Random Forests (RF) that were processed and applied to the statistical software R. Levels of thematic accuracy were obtained for each these methods and the comparison was made between them, leading to the conclusion that for the study area, the Random Forests can provide a better extraction of thematic information of land cover which is seen in the values obtained and the results visually, although the classification obtained with Decision Trees is also good.

Last modified: 2021-03-18 21:05:17