A Novel Approach for Resource Estimation of Highly Skewed Gold Using Machine Learning Algorithms

With the complicated geology of vein deposits, their irregular and extremely skewed grade distribution, confined nature gold, there is a propensity to overestimate or underestimate ore grade. As result, numerous estimation approaches for mineral resources have been developed. It was investigated in this study by using five machine learning algorithms estimate highly gold data vein-type at Quartz Ridge region, including Gaussian Process Regression (GPR), Support Vector (SVR), Decision Tree Ensemble (DTE), Fully Connected Neural Network (FCNN), K-Nearest Neighbors (K-NN). The accuracy MLA compared that geostatistical approaches, such as ordinary indicator kriging. Significant improvements were made during preprocessing splitting, ensuring estimated accurately. preprocessed with two normalization methods (z-score logarithmic) enhance network training performance minimize substantial differences dataset’s variable ranges on predictions. samples divided into equal subsets an integrated segmentation approach based Marine Predators Algorithm (MPA). ranking shows GPR logarithmic most efficient method estimating grade, far outperforming kriging techniques. In study, key producing successful more than just technique. also has do how are processed split.