Optimized Weighted Ensemble Approach for Enhancing Gold Mineralization Prediction

The economic value of a mineral resource is highly dependent on the accuracy grade estimations. Accurate predictions grades can help businesses decide whether to invest in mining project and optimize operations maximize resource. Conventional methods predicting gold resources are both costly time-consuming. However, advances machine learning processing power making it possible for estimation become more efficient effective. This work introduces novel approach distribution within deposit. integrates optimization techniques. Specifically, authors propose an that random forest (RF) k-nearest neighbor (kNN) algorithms with marine predators algorithm (MPA). RFKNN_MPA uses log normalization reduce impact extreme values improve models. Data segmentation MPA used create statistically equivalent subsets dataset use training testing. Drill hole locations rock types each model. suggested technique’s performance indices superior others, higher R-squared coefficient 59.7%, R-value 77%, lower MSE RMSE 0.17 0.44, respectively. outperforms geostatistical conventional machine-learning techniques estimating orebody grades. introduced offers solution problem practical applications sector.