Advancing Mineral Resource Estimation with Machine Learning: A Case Study from Ransko.

Exploration drilling is a crucial yet expensive process for gaining insights into subsurface environments. With the rising demand for critical minerals needed for the green energy transition, the number of exploration projects has significantly increased. Traditional geostatistical methods are commonly used for mineral resource estimation, but they often depend on dense and extensive datasets, making them challenging for small-scale explorations and environmentally sensitive areas.  This study explores the use of machine learning (ML) techniques, specifically Extreme Gradient Boosting and Random Forest, to improve mineral resource estimation in the Ransko region. ML methods present a groundbreaking approach by predicting target variables in unsampled locations using minimal and distant data, effectively reducing environmental impact and exploration costs. Additionally, ML can incorporate geological interpretations and account for spatial continuity, enhancing the quality of estimates and leading to more efficient and sustainable mineral exploration practices.