Land use and land cover classification using GEE and machine learning algorithms: a case study of Vaijapur Tehsil

The monitoring and analysis of land surface changes are very challenging for researchers in terms of time and the cost of software tool licenses. Now, it has become easier to accomplish with the help of modern tools like Google Earth Engine (GEE) and built-in Machine Learning (ML) algorithms. GEE is a platform for analysing and interpreting satellite datasets in the cloud, which saves time and storage space for researchers. In the present research study, we have analysed the Land Use Land Cover (LULC) of Vaijapur taluka, which is undergoing a rapid shift from rural to urban. The study compares the LULC classification by three supervised ML algorithms, Random Forest (RF), k- Nearest-Neighbour (k-NN), and Gradient Tree Boost (GTB). We have obtained the overall accuracy of RF (92%) is better than that of GTB (88 %) and k-NN (86.67%). The research study observed that the kappa accuracy of RF (0.89) is improved than GTB (0.83) and k-NN (0.81).