Study of the expansion of urban space using the CART method

Changes in land use and urban development are an important factor in the study of urban expansion. Machine learning classifiers are increasingly being developed for land-use and land-cover (LULC) mapping, especially urban space. This comparison shows CART, a simple binary decision tree classifier that recursively splits nodes until a predefined threshold value is reached, was used in global LULC mapping. The classification accuracy and fast performance of CART make it one of the widely used LULC classifiers, as demonstrated in this comparison