CBERS-02 Remote Sensing Data Mining Using Decision Tree Algorithm

Abstract

In recent years, decision tree algorithms have been successfully used for land cover classification from remote sensing data. In this paper, CART (classification and regression trees) and C5.0 decision tree algorithms were used to CBERS-02 remote sensing data. Firstly, the remote sensing data was transformed using the Principal Component Analysis (PCA) and multiple-band algorithm. Then, the training data was collected from the combining total 20 processed bands. Finally, the decision tree was constructed by CART and C5.0 algorithm respectively. Comparing two results, the most important variables are clearly band, band and band. The depth of the CART tree is only two with the relative high accuracy. The classification outcome was calculated by CART tree. In order to validate the classification accuracy of CART tree, the Confusion Matrices was generated by the ground truth data collected using visual interpretation and the field survey and the kappa coefficient is 0.95.