Change Detection Using Data Stacking and Decision Tree Techniques in Puer-Simao Counties of Yunnan Province, China

Classification performing expert system Decision Tree and Ctree model were used to assess the land use cover change (LULCC). The Classification was done in Excel CTree program and the land cover change analyzing in ENVI software tools. Tree based Classification Model consisted of four steps (i) input the data, (ii) model processing, (iii) Analyzing of tree built and (iv) rules generation. The logic contained in the decision rules derived by that program was used to build a decision tree classifier with ENVI"s interactive decision tree tool. The premise of a rule-based modeling approach is that distinct land cover types are associated with different ranges of environmental and spectral gradients, and that "rules" can be drawn from spectral and ancillary modeling layers to correctly identify the spatial distribution of land cover classes. Rules are normally expressed in the form of one or more "IF condition THEN action" statements. All the three classification dates achieved high overall accuracies of 94, 97 and 92% for 1999, 2002 and 2005 respectively. The integration of Ctree program and Decision Tree was a good opportunity in land use land cover change detection in Puer-Simao counties, it can be implemented in others similar studies. (Diallo Y, Diarra ST, Sagara B, Wen X, Xu Y, Bokhari, A A, Hu G. Change Detection Using Data Stacking and Decision Tree Techniques in Puer-Simao Counties of Yunnan Province, China. Researcher 2015;7(3):1-12). (ISSN: 1553-9865). http://www.sciencepub.net/researcher. 1