Scientists at the Trofimuk Institute Develop Neural Network for Mineral Exploration
The algorithm can determine the depth at which resources are located.
Scientists at the Trofimuk Institute of Petroleum Geology and Geophysics of the Siberian Branch of the Russian Academy of Sciences have developed a neural network for analyzing electromagnetic sounding data from the subsurface. The technology can quickly determine the depth of highly conductive resources, including certain types of ores.
The model was trained on a dataset containing 125 electromagnetic force (EMF) signal patterns generated by conductive, layer-shaped objects in a weakly conductive environment. The algorithm can determine the depth of the target objects with an error of no more than 15%.
The researchers plan to improve the forecasting accuracy by continuing to train the neural network on a larger and more diverse dataset.
