Rock Classification in a Vanadiferous Titanomagnetite Deposit Based on Supervised Machine Learning

As the potential locations of undiscovered ore deposits become deeper, a technique for predicting promising areas in subsurface media has necessary. Geoscience data on wide range underground can be obtained through geophysical field exploration, but integration and interpretation multi-geophysical are difficult because differences spatial resolution. We developed rock classifier that predict vanadiferous titanomagnetite from using supervised machine learning. Vanadiferous ores main source vanadium, which used as large-scale energy storage system. Model training was conducted samples drilling cores, density criterion labeling. employed support vector machine, random forest, extreme gradient boosting, LightGBM, deep neural network learning, accuracy all methods 0.95 or greater. applied trained models to three-dimensional body locations. These candidate regions were distributed northeast survey area, some classified verified geological map.