USING GIS DATA AND MACHINE LEARNING FOR MINERAL MAPPING. STUDY CASE, BOU SKOUR EASTERN ANTI-ATLAS, MOROCCO
Keywords: Deposit, Prediction, Geology, Machine learning, Mineral mapping, Random Forest, Artificial Neural Network
Abstract. The continued demand for mineral deposits in recent years has led exploration geologists for each stage of mineral exploration; find more effective and innovative ways of processing different data types. The use of Geographic Information Systems (GIS) allows various features, such as elevation, slope, tectonic structures, lithological units and indicator minerals of Bou Skour region, Eastern Anti-Atlas, Morocco to be mapped making targeted mining decisions easier. In this paper, a methodology was developed to enable the automated mapping of mineral using machine learning methods such Random Forest (RF) and Artificial Neural Network (ANN) achieves approximately 98% classification accuracy on a single Intel® Core™ i5-5300U CPU core with 16GB of memory, and come up with predictive maps representing the probable potentially mineralized areas.