Automated mapping of metallotects with the use of deep machine learning methods

This article provides a brief overview of the theory of neural network methods of pattern recognition. It considers the features of using deep machine learning methods in interpreting complex geological and geophysical data and the possibilities of using them to solve forecasting problems. A method for neural network forecasting of ore mineralization, based on a priori physical-geological modeling by automated mapping of geological ore control factors, is proposed. The method is demonstrated on the example of identifying gold ore deposits in the Central Kolyma region.