The weak classifier ensemble algorithms based on the decision tree model, mainly include bagging (e.g., fandom forest-RF) and boosting (e.g., gradient boosting decision tree, eXtreme gradient boosting), the former reduces the variance for the overall generalization error reduction while the latter focuses on reducing the overall bias to that end. Because of its straightforward idea, it is prevalent in MPM (mineral prospectivity mapping). However, an inevitable problem in the application of such methods is the hyperparameters tuning which is a laborious and time-consuming task. The selection of hyperparameters suitable for a specific task is worth investigating. In this paper, a tree Parzen estimator-based GBDT (gradient boosting decision tree) model (TPE-GBDT) was introduced for hyperparameters tuning (e.g., loss criterion, n_estimators, learning_rate, max_features, subsample, max_depth, min_impurity_decrease). Then, the geological data of the gold deposit in the Xiong ‘ershan area was used to create training data for MPM and to compare the TPE-GBDT and random search-GBDT training results. Results showed that the TPE-GBDT model can obtain higher accuracy than random search-GBDT in a shorter time for the same parameter space, which proves that this algorithm is superior to random search in principle and more suitable for complex hyperparametric tuning. Subsequently, the validation measures, five-fold cross-validation, confusion matrix and success rate curves were employed to evaluate the overall performance of the hyperparameter optimization models. The results showed good scores for the predictive models. Finally, according to the maximum Youden index as the threshold to divide metallogenic potential areas and non-prospective areas, the high metallogenic prospect area (accounts for 10.22% of the total study area) derived by the TPE-GBDT model contained > 90% of the known deposits and provided a preferred range for future exploration work.

The demand for graphite resources has been increasing due to its extensive use. Graphite deposits in China are mainly distributed in Heilongjiang, Innner Mogolia, Sichuan, Shanxi and Shandong, characterized by "one old & one new" and "many in east & few in west". There are mainly three genetic types, including regional metamorphic type, contact metamorphic type and hydrothermal type. Here we provide a summary of the metallogenic regularity and 15 metallogenic belts in China based on the study of the geology of national graphite deposits. In recent years, great progress has been made in the research of genesis, mineral exploration and evaluation on graphite deposits, which indicate great potential for graphite resource in China. The authors suggest that the research on geochronology, ore sources, the genesis of crystalline graphite and exploration for hydrothermal graphite in China should be heightened.

Large porphyry Cu and epithermal Au deposits tend to form in distinct tectonic, porphyry and high-sulfidation epithermal deposits in compressional settings, and low-sulfidation epithermal deposits in extensional settings. Given that the analysis of the shallow metallogenic dynamic processes at the upper-crust scale is insufficient, especially the ore-bearing fracture formation and fluid-focusing mechanism around the mineralizing magmatic intrusion under different tectonic backgrounds, we aimed to study how tectonic settings influence fracture formation and fluid hydrodynamics in and around a hot intrusion. We developed a finite element model coupling thermal-hydrological-mechanical processes to simulate the fracture formation, evolution of fluid velocities, and accumulation of water-rock interactions. The model results show that tectonic compression increases the degree of fracturing, hydrothermal fluid velocities, and water-rock interaction within and laterally around the intrusion; tectonic extension enhances fracturing, hydrothermal fluid velocities, and water-rock interaction at shallow depth. These results confirm that tectonic compression may promote the formation of porphyry Cu deposits, while tectonic extension may promote the formation of shallow hydrothermal deposits. Our model explains the effects of tectonic activity on fracture formation and fluid flow around hot magmatic intrusions in upper crust and deepens our understanding of the relationship between tectonic activity and deposit formation there.

In order to achieve a breakthrough in the exploration of uranium orebodies in the Changjiang area of Guangdong province in China, the geo-electrochemical exploration method is used for mineral resource prediction. The logarithmically and isometric log-ratio (ilr) transformations are applied to the geo-electrochemical data in this study area to extract geochemical anomalies. The relationship between element associations and mineralization is revealed through descriptive statistical analysis and further biplot analysis. Then, the energy spectrum density–area fractal model (S-A model) is used to identify geochemical backgrounds and anomalies. The results show that: (i) the logarithmically and ilr-transformed data are more uniform, and they more or less obey the rules of normal distribution; (ii) the biplot shows that the ilr-transformed data eliminates the closure effect, and the robust principal component analysis (RPCA) has a better indicative significance for element associations—PC1 reveals the mineralization element association dominated by U and the multiple periods of hydrothermal activity; (iii) the S-A method could extract the local anomalies from different geochemical backgrounds, which indicates mineralization is more reliable. Finally, four favorable prospecting targets are delineated based on the geological and geochemical indicators.