(R² = 0.7764), indicating its ability to capture a substantial portion of spatial variability in mineralization occurrence. In binary classification, SVR reached a precision of 0.9800 and the highest F1-score (0.6667), reflecting an effective balance between precision and recall under a conservative threshold (0.7200). Its recall (0.5000) indicates that the model prioritizes certainty over exhaustiveness, favoring high-confidence target delineation while minimizing false positives, and suggesting suitability for cost-sensitive exploration contexts where over-drilling is a critical risk. Despite higher absolute regression errors (MAE and RMSE) relative to other models, attributable to broader score dispersion, SVR achieved superior classification reliability under optimized thresholding. Extreme Gradient Boosting (XGB) demonstrated the strongest discriminative capacity (AUC = 0.9989) and complete recall (1.0000), successfully classifying all mineralized cells in the validation dataset. However, its precision (0.0036) was low under the selected classification threshold (0.0040), indicating substantial false positives. This behavior reflects strong ranking performance but sensitivity to threshold selection in strongly imbalanced class conditions. The limited number of known deposits relative to non-mineralized cells creates strong class imbalance, which may inflate AUC and recall while increasing false positives, potentially raising exploration validation costs. The heuristic Empirical Weighted Mean (Wmd) exhibited negligible explanatory power (R² = 0.0045) and extremely low precision (0.0017), highlighting the limitations of linear, expertdriven combinations in modeling complex geological interactions. Similarly, while KNN and LGB achieved moderate AUC values—suggesting a latent ability to rank prospective zones—they failed to reach meaningful binary classification performance (F1-score = 0.0000). This discrepancy suggests that while non-linear machine learning models are better equipped to capture the intricate interactions among hydrothermal signatures and structural controls, they require careful threshold optimization to overcome the inherent class imbalance in mineral prospectivity mapping. Consequently, the most reliable prospectivity maps for operational deployment were generated by SVR (conservative targeting) and XGB (exhaustive screening). These maps not only successfully delineated known mineralized zones but also identified new, previously unrecognized high-potential areas, validating the utility of this machine learning-based approach for mineral exploration. Table 1 – Metrics and Hyperparameters associated with the predictive methods Paramet er Wmd RF GAM XGB LGB KNN SVR R² 0.0045 0.1001 0.0224 0.3750 0.0823 0.0763 0.7764 MAE 0.0373 0.0007 0.0014 0.0006 0.0010 0.0006 0.1498 RMSE 0.0569 0.0138 0.0144 0.0118 0.0139 0.0139 0.1499
RkJQdWJsaXNoZXIy MTM0Mzk2