Semi-automated geological mapping and target generation from geochemical and magnetic data in Halkidiki region, Greece

This study focuses on using multivariate analyses to generate semi-automated geological maps and exploration targets associated with porphyry Au-Cu mineralization within the Kassandra mining district, Greece. We use principal component analysis (PCA) self-organizing (SOM) reveal variations in geochemical magnetic signatures input datasets. visualize results as pseudo-geological reflecting processes their end products. In specific, we utilize potential of these two methods through an integrated interpretation comparison results. test validity unsupervised PCA- SOM-derived lithological prospectivity models by comparing them existing observations interpretations. The this investigation show that both PCA SOM are able reproduce key features area, but more importantly, also provide useful information can be used recognize prospective units from previously unknown locations.