Probabilistic neural networks and fractal method applied to mineral potential mapping in Luanchuan region, Henan Province, China

This paper presents an application of probabilistic neural networks (PNN) to integrated analysis multi-mineral anomalies caused by geological information (geology, geophysics, geochemistry, and remote sensing) and to map the 1:25000 scale potential for Molybdenum polymetallic Pb-Zn-Ag mine targets with in Luanchuan region, Hennan Province. On the one hand, according to geological anomaly theory, the use of GIS technologies for the study area of geological anomaly information extraction, Mo ore-forming elements of the multi-fractal anomaly delineation, ETM+ remote sensing data, hydroxyl and iron staining alteration of information extraction, gravity and high magnetic anomaly deep geological body inversion; the other hand, the use of PNN method (via the probability density function (non-linear Gauss transform function) for complex non-linear classification), to carry out the study area to study geoanomaly associated with mineralization (variable) integrated analysis and metallogenic prediction. The results show that PNN method combined with fractal analysis can not only integrate the study area Pb-Zn-Ag-Mo polymetallic mines pluralistic, multi-scale and multi-types of geoanomaly associated with mineralization, but also carved out of the study area of molybdenum (tungsten), and Pb-Zn-Ag are two types of mineralization favorable target areas, and this work provides a scientific basis for the deployment of mineral exploration projects in the study area.