Predictive Deep Neural Network Model of Doukkala Coastal Domain Land Use with Remote Sensing Data
摘要
Remote Sensing is a constantly evolving field, and contain an enormous potential for analyzing satellite images. Although, in order to deal with many geoscience phenomena such as Land Use prediction, it requires opening up to new technologies for data analysis and knowledge discovery. We worked in this paper on the development of a rigorous Artificial Neural Network model, applied to our Data Set elaborated especially to our project, and composed of three indices: Normalized difference vegetation index (NDVI), Normalized difference water index (NDWI) and Dry built-up index (DBI). The goal is the support of decision-makers and managers, in order to decide on useful investments for the region.