Mineral Prospectivity Mapping Research Articles

Analysing multi-index overlay and fuzzy logic models for lode-gold prospectivity mapping in the Ahafo gold district – Southwestern Ghana

Mineral prospectivity mapping (MPM) is a multi-criteria decision-making process that identifies and prioritises potential zones for mineral exploration. Multi-index overlay and fuzzy logic models were employed in this paper to conduct GIS-based MPM of the Ahafo gold district, which is located on the northwestern margin of the Sefwi granitegreenstone belt in Ghana. Mineralisation in the area is structurally controlled with a high degree of hydrothermal alteration. Therefore, the mineral systems approach was used to translate the understanding of orogenic gold hydrothermal system into mappable exploration criteria, leading to the generation of nine evidential layers. These layers represented mappable spatial proxies for active fluid pathways (i.e. proximity to NEstriking basement faults: D2 to D3), physical traps (i.e. proximity to N- to NNE-strike sinistral fault: D4; proximity to litho-contacts; proximity to the intersection of N- to NNE- and NE-strike faults; density of N- to NNE- and NE-strike faults; litho-competency contrast), and chemical traps (i.e. proximity to Fe-rich dykes; litho-reactivity contrast; proximity to hydrothermal alteration zone), all of which are critical for hydrothermal fluid migration, focusing, and gold deposition. Weights and confidence factors were assigned to the evidential layers in the fuzzy logic approach based on experts' knowledge of how the layers are relevant to orogenic gold mineralising processes. In contrast, weights were assigned to layers in the multi-index overlay model based on prediction-area (P-A) plots and normalised densities, limiting the subjectivity of assigning weights to the layers. In order to delineate potential zones for lode-gold mineralisation in the study area, the weighted evidential layers were combined to create multi-index overlay and fuzzy prospectivity maps. The continuous prospective scores of the prospectivity maps were discretised and classified using a concentration-area (C-A) fractal model. The resulting discretised prospectivity maps based on the location of 18 known lode-gold occurrences (i.e. deposits and economic prospects) were further evaluated using P-A plots. The discretised multi-index overlay prospectivity map predicted 76% of known lode-gold occurrences, with 24% of the study area as prospective, whereas the discretised fuzzy prospectivity map predicted 74% of known lode-gold occurrences, with 26% of the study area as prospective. The high potential zones identified on the prospectivity maps were spatially located in areas with structural complexity along: (i) NE-SW structural trends (i.e. 40°–50°); and (ii) secondary structural trends in the NS to NNE-SSW direction (i.e. 0°–10° and 15°–25°). These structural trends control lode-gold mineralisation in the Ahafo gold district..

Machine Learning Methods for Quantifying Uncertainty in Prospectivity Mapping of Magmatic-Hydrothermal Gold Deposits: A Case Study from Juruena Mineral Province, Northern Mato Grosso, Brazil

Mineral prospectivity mapping (MPM), like other geoscience fields, is subject to a variety of uncertainties. When data about unfavorable sites to find deposits (i.e., drill intersections to barren rocks) is lacking in MPM using machine learning (ML) methods, the synthetic generation of negative datasets is required. As a result, techniques for selecting point locations to represent negative examples must be employed. Several approaches have been proposed in the past; however, one can never be certain that the points chosen are true negatives or, at the very least, optimal for training. As a consequence, methodologies that account for the uncertainty of the generation of negative datasets in MPM are needed. In this paper, we compare two criteria for selecting negative examples and quantify the uncertainty associated with this process by generating 400 potential maps for each of the three ML methods utilized (200 maps for each criterion), which include random forest (RF), support vector machine (SVM), and k-nearest neighbors (KNC). The results showed that applying a geological constraint to the creation of negative datasets reduced prediction uncertainty and improved overall model performance but produced larger areas of very high probability (i.e., >0.9) when compared to using only the spatial distribution of known deposits and occurrences as a constraint. SHAP values were used to find approximations for the importance of features in nonlinear methods, and kernel density estimations were used to examine how they varied depending on the negative dataset used to train the ML models. Prospectivity models for magmatic-hydrothermal gold deposits were generated using data from the shuttle radar terrain mission, gamma-ray, magnetic lineaments, and proximity to dykes. The Juruena Mineral Province, situated in Northern Mato Grosso, Brazil, represented the case study for this work.

Quantifying Uncertainties Linked to the Diversity of Mathematical Frameworks in Knowledge-Driven Mineral Prospectivity Mapping

Quantifying Uncertainties Linked to the Diversity of Mathematical Frameworks in Knowledge-Driven Mineral Prospectivity Mapping

Unlabeled Sample Selection for Mineral Prospectivity Mapping by Semi-supervised Support Vector Machine

Unlabeled Sample Selection for Mineral Prospectivity Mapping by Semi-supervised Support Vector Machine

Metallogenic models as the key to successful exploration — a review and trends

Metallogeny is the science of ore and mineral deposit formation in geological space and time. Metallogeny is interdisciplinary by nature, comprising elements of natural science disciplines such as planetology to solid state physics and chemistry, and volcanology. It is the experimental forefront of research and bold thinking, based on an ever-growing foundation of solid knowledge. Therefore, metallogeny is not a closed system of knowledge but a fast-growing assemblage of structured and unstructured information in perpetual flux. This paper intends to review its current state and trends. The latter may introduce speculation and fuzziness. Metallogeny has existed for over 100 years as a branch of Earth Science. From the discovery of plate tectonics (ca. 1950) to the end of the last century, metallogeny passed through a worldwide phase of formally published ‘metallogenetic’ maps. In the last decades, a rapidly growing number of scientists, digitization and splendid new tools fundamentally boosted research. More innovations may be expected by the growing use of an evolving systematic ‘Geodata Science’ for metallogenic research by an increasingly global human talent pool. Future requirements for metallic and mineral raw materials, especially the critical natural elements and compounds that are needed for the nascent carbon-free economy, already drive activities on stock markets and in the resource industry. State geological surveys, academia and private companies embrace the challenges. The new age requires intensified metallogenic backing. In this paper, principles of metallogeny are recalled concerning concepts and terms. A metallogenic classification of ore and mineral deposits is proposed, and the intimate relations of metallogenesis with geodynamics are sketched (ancient lid tectonics and modern plate tectonics). Metallogenic models assemble a great diversity of data that allow an ever better understanding of ore formation, foremost by illuminating the geological source-to-trap migration of ore metals, the petrogenetic and geodynamic–tectonic setting, the spatial architecture of ore deposits and the nature and precise timing of involved processes. Applied metallogeny allows companies to choose strategy and tactics for exploration investment and for planning the work. Based on comprehensive metallogenic knowledge, mineral system analysis (MSA) selects those elements of complex metallogenic models, which are detectable and can guide exploration in order to support applications such as mineral prospectivity mapping, mineral potential evaluation and targeting of detailed investigations. MSA founded on metallogenic models can be applied across whole continents, or at the scale of regional greenfield search, or in brownfields at district to camp scale. By delivering the fundamental keys for MSA, supported by unceasing innovative research, the stream of new metallogenic insights is essential for improving endowment estimates and for successful exploration.

Developments in Quantitative Assessment and Modeling of Mineral Resource Potential: An Overview

The special issue entitled “Developments in Quantitative Assessment and Modeling of Mineral Resource Potential” is composed of 17 papers that cover a diverse range of approaches to mineral resource assessment, including mainly multivariate statistical analysis, fractal and multifractal modeling, geostatistical modeling, machine learning, mathematical morphology and 3D mineral exploration. This introductory article first provides an overview of the developments and existing methods in quantitative assessment of mineral resources. Then, brief introductions of each of the 17 papers are provided. These papers can be grouped into three themes: (1) multifractal theory for mineral resource assessment; (2) machine learning for mineral prospectivity mapping; and (3) GIS-based 3D modeling for mineral exploration. The 17 papers either proposed novel methods or demonstrated innovative applications of existing methods.

Identification of Radioactive Mineralized Lithology and Mineral Prospectivity Mapping Based on Remote Sensing in High-Latitude Regions: A Case Study on the Narsaq Region of Greenland

The harsh environment of high-latitude areas with large amounts of snow and ice cover makes it difficult to carry out full geological field surveys. Uranium resources are abundant within the Ilimaussaq Complex in the Narsaq region of Greenland, where the uranium ore body is strictly controlled by the Lujavrite formation, which is the main ore-bearing rock in the complex rock mass. Further, large aggregations of radioactive minerals appear as thermal anomalies on remote sensing thermal infrared imagery, which is indicative of deposits of highly radioactive elements. Using a weight-of-evidence analysis method that combines machine-learned lithological classification information with information on surface temperature thermal anomalies, the prediction of radioactive element-bearing deposits at high latitudes was carried out. Through the use of Worldview-2 (WV-2) remote sensing images, support vector machine algorithms based on texture features and topographic features were used to identify Lujavrite. In addition, the distribution of thermal anomalies associated with radioactive elements was inverted using Landsat 8 TIRS thermal infrared data. From the results, it was found that the overall accuracy of the SVM algorithm-based lithology mapping was 89.57%. The surface temperature thermal anomaly had a Spearman correlation coefficient of 0.63 with the total airborne measured uranium gamma radiation. The lithological classification information was integrated with surface temperature thermal anomalies and other multi-source remote sensing mineralization elements to calculate mineralization-favorable areas through a weight-of-evidence model, with high-value mineralization probability areas being spatially consistent with known mineralization areas. In conclusion, a multifaceted remote sensing information finding method, focusing on surface temperature thermal anomalies in high-latitude areas, provides guidance and has reference value for the exploration of potential mineralization areas for deposits containing radioactive elements.

Hybrid outranking of geospatial data: Multi attributive ideal-real comparative analysis and combined compromise solution

Identifying mineralization areas with an acceptable level of reliability is a complex matter demanding the implementation of supervised methods for mapping mineralization exploration aims. In this study, further collecting all available exploratory information and data of the region of study, their good processing and analysis, and then integrating them with an appropriate method, multi-criteria decision-making (MCDM) approaches were utilized to outranking porphyry Cu mineralization areas. This paper describes an acceptable procedure for obtaining evidential layers for recognizing porphyry Cu mineralization, assigning weight to the layers, and combining them. For this aim, first, a data-driven multi-class index overlay (DMIO) approach is applied as it is a proven method with proper results in mineral prospectivity mapping (MPM) to integrate evidential layers (criteria and sub-criteria emanated from geoscience data including geological, geochemical, remote sensing, and geophysical datasets). After that, two recent MCDM models, combined compromise solution (CoCoSo) and multi attributive ideal-real comparative analysis (MAIRCA), were used to prioritize the porphyry Cu potential areas producing MPM. Concentration-area (C–A) fractal method and prediction-area (P–A) plots by considering the areas of occurrence of the known mineralization and normalized density (Nd) as traditional methods were applied for weighting and integration evidential layers and evaluation of the final maps in a data-driven manner. Consequently, verifying the verisimilitude of the MAIRCA prospectivity map (Nd = 3.17) is similar to the DMIO map in delimiting the priority areas. The results of the CoCoSo model (Nd = 2.7) show this methodology was also successfully applied in mapping the porphyry Cu mineralization areas in the study.

Ensemble learning models with a Bayesian optimization algorithm for mineral prospectivity mapping

Machine learning algorithms have been widely applied in mineral prospectivity mapping (MPM). In this study, we implemented ensemble learning of extreme gradient boosting (XGBoost) and random forest (RF) models to create MPM for magmatic hydrothermal tin polymetallic deposits in Xianghualing District, southern Hunan Province, China. Machine-learning models often require careful adjustment of the learning parameters and model hyperparameters for optimal global performance. However, parameter tuning often entails tedious calculations and sufficient expert experience, which is a time-consuming and labor-intensive process. To obtain the global optimal performance of the XGBoost and RF models, a Bayesian optimization algorithm (BOA) was employed with the aid of 5-fold cross validation to search for the most appropriate hyperparameters of the XGBoost and RF models. After the Bayesian optimization, the AUC values of both models were significantly improved, indicating that the BOA is a powerful optimization tool. The optimization results provide a reference for the empirical hyperparameter setting of ensemble learning models. Through a comparative study, the XGBoost model was shown to be superior to the RF model in terms of accuracy, precision, recall, F1 score, and kappa coefficient. In addition, the receiver operating characteristic curves and prediction–area curves showed that the XGBoost model outperformed the RF model, indicating that the XGBoost model had better prediction ability and stability in the case area. In this study, the XGBoost model shows great potential for MPM, offering a significant improvement over the BOA method.

Three-Dimensional Mineral Prospectivity Mapping by XGBoost Modeling: A Case Study of the Lannigou Gold Deposit, China

Three-Dimensional Mineral Prospectivity Mapping by XGBoost Modeling: A Case Study of the Lannigou Gold Deposit, China

A Geologically Constrained Variational Autoencoder for Mineral Prospectivity Mapping

A Geologically Constrained Variational Autoencoder for Mineral Prospectivity Mapping

Toward a mineral potential map of central Aceh, North Sumatra: contributions from spectrum-area and local singularity fractal analysis of stream sediment geochemistry and its relationships to magmatic-hydrothermal Cu and Au mineralisation

ABSTRACT A reassessment of the geochemistry of stream sediments collected by the British-led North Sumatra Project from 1975 to 1980 in central Aceh with spectrum–area and local singularity fractal techniques refines previously determined areas with anomalous Cu and delineates As anomalies (proxy for Au) that correspond to porphyry Cu and epithermal Au–Ag deposits and prospects discovered in the 1980s and 1990s. Additionally, the application of these methods to stream sediment samples collected by an Australian company in the mid-1990s from part of central Aceh identifies Au, As, and Cu anomalies that are aligned with regional geochemical trends. Integrating singularity metal-enriched areas and the distribution of mineral occurrences and structures reveals northerly and easterly tectono-magmatic corridors controlling epithermal Au–Ag and porphyry Cu mineralisation. The above results attest to the usefulness of legacy data for first-pass mineral assessments and for providing evidence layers for a future mineral prospectivity map.

Data-driven multi-index overlay gold prospectivity mapping using geophysical and remote sensing datasets

The comparative relevance of each geospatial component of mineralization differs from one geological terrane to the other because various sought-after mineral deposit-types synonymously differ in different geological terranes. Hence, the possibility of employing a conceptual model to obtain a relationship or a quantitative function between various geospatial features (evidential layers) with respect to the mineral being sought is laudable, though these features may not necessarily have a generically related effect with the mineral being sought. As a consequence, there is the need to employ a technique that has the capacity to recognize the efficient and inefficient geospatial indicators of the mineral deposit-type being sought. In view of this, this study employed the logistic function, concentration-area fractal model and the prediction-area (P-A) plot to transform and discretize the continuous value of each evidential layer as well as generating intersection points of prediction rate indicators that are essential in obtaining the normalized densities, which were subsequently employed in generating the objective weight for each evidential layer in a data-driven way. The P-A and the normalized density techniques employed were vital in recognizing the indicator and non-indicator criteria. The results obtained acknowledged the potassium concentration layer as a non-indicator of gold mineralization within the study area and subsequently recognized the hydroxyl bearing mineral concentration layer as the most plausible indicator criteria among the six evidential layers (lineament density, iron concentration, hydroxyl concentration, gravity anomaly, magnetic anomaly and potassium concentration) employed in this study. These five indicator criteria were integrated to generate a mineral prospectivity map (MPM) over the study area based on the data-driven multi-index overlay approach adopted. The prediction rate for each of the 6 evidential layers (5 of which were the indicator criteria) as well as the MPM produced indicates that, the generation of objective weights in a data-driven manner via normalized density enhances the predicting ability of the MPM produced in comparison with the individual evidential layers.

Applications of data augmentation in mineral prospectivity prediction based on convolutional neural networks

The supervised deep learning methods applied in mineral prospectivity mapping usually need sufficient samples for training models. However, mineralization is a rare event. Insufficient known mineral deposits cannot meet the sample requirement of supervised learning methods, resulting in lower predictive accuracies and poor generalization abilities. For the purpose of solving this issue, this paper adopted a data augmentation method to make mineral prospectivity prediction of gold deposit in the Fengxian Region, China. This data augmentation method adopted cropping operations to generate sufficient training samples without changing spatial directions of geological data. Meanwhile, this paper utilized the continuous buffer distance method to quantify faults and anticline axes, overcoming the loss of geological information caused by using the discrete buffer distance mode. To prove the effectiveness of the data augmentation method, this paper utilized three different convolutional neural networks (LeNet, AlexNet, and VggNet) to extract relationships between multisource ore-indicating factors and mineral deposits. In addition, this paper discussed effects of different parameters on predictive performances. According to series of comparisons, the LeNet model outperformed other models, achieving superior values of accuracy (91.38%), Kappa coefficient (0.8119), and AUC (0.958). Moreover, the LeNet model successfully caught 81.8% of known gold deposits within 18.6% of the study area. The delineated high potential areas offer intuitive guides for exploring more gold deposits in the Fengxian region. The proposed data augmentation method is available for mineral prospectivity modeling by supervised deep learning methods for the areas of lower exploration degrees. For mineral prospectivity modeling based on convolutional neural networks, utilizing the continuous buffer distance to transform faults and anticline axes into predictor variables of the image form is conducive to improve the predictive performance than utilizing the discrete buffer distance.

Mineral prospectivity mapping based on wavelet neural network and Monte Carlo simulations in the Nanling W-Sn metallogenic province

The Nanling Range in South China is endowed with abundant W-Sn and other important rare metal resources associated with granitic intrusions, but the rate of new major mineral discoveries has been drastically diminished over the last decades. Data-driven mineral prospectivity mapping (MPM) using machine learning is emerging as a powerful tool in support of mineral exploration targeting at regional or camp scale. However, MPM based on supervised learning faces some general problems, notably including subtle information obscured in exploration data and imbalanced training samples. In this paper, we proposed a novel ensemble scheme for MPM using wavelet neural network (WNN) and Monte Carlo simulation (MCs) to address the forementioned issues. Specifically, the purpose of using WNN based on multiscale wavelet analysis is to capture weak or hidden (geophysical and geochemical) information caused by concealed ore deposits, while an undersampling MCs (UMCs) algorithm is employed to assess uncertainties caused by data imbalance and lessen its impact on MPM. Subtle (weak and mixing) information related to concealed mineralization, including geophysical and geochemical anomalies, were extracted using wavelet analysis at first, and then a dozen of predictive evidences was integrated based on WNN using UMCs for producing a W-Sn perspectivity map. The results suggest that WNN has a higher prediction accuracy (∼90%) in comparison to the artificial neural network (ANN) (∼85%). Moreover, the improved AUC value (∼0.9) of WNN based on UMCs compared to the classic ANN (∼0.82) and WNN (∼0.85) suggests that the proposed ensemble algorithm improves the geological generalization of machine learning. Revisiting WNN predictions by uncertainties produced more sharply focused exploration targets of W-Sn deposits (covering ∼5% of the study area). The resulting predictive map provides important clues of W-Sn deposit occurrences which could guide and stimulate future mineral exploration in the Nanling Range.

A Comparative Study of Convolutional Neural Networks and Conventional Machine Learning Models for Lithological Mapping Using Remote Sensing Data

Lithological mapping is a critical aspect of geological mapping that can be useful in studying the mineralization potential of a region and has implications for mineral prospectivity mapping. This is a challenging task if performed manually, particularly in highly remote areas that require a large number of participants and resources. The combination of machine learning (ML) methods and remote sensing data can provide a quick, low-cost, and accurate approach for mapping lithological units. This study used deep learning via convolutional neural networks and conventional ML methods involving support vector machines and multilayer perceptron to map lithological units of a mineral-rich area in the southeast of Iran. Moreover, we used and compared the efficiency of three different types of multispectral remote-sensing data, including Landsat 8 operational land imager (OLI), advanced spaceborne thermal emission and reflection radiometer (ASTER), and Sentinel-2. The results show that CNNs and conventional ML methods effectively use the respective remote-sensing data in generating an accurate lithological map of the study area. However, the combination of CNNs and ASTER data provides the best performance and the highest accuracy and adaptability with field observations and laboratory analysis results so that almost all the test data are predicted correctly. The framework proposed in this study can be helpful for exploration geologists to create accurate lithological maps in other regions by using various remote-sensing data at a low cost.

Application of maximum entropy for mineral prospectivity mapping in heavily vegetated areas of Greater Kurile Chain with Landsat 8 data

Exploration for strategic mineral resources, like precious metals, in remote areas of the Greater Kurile Chain is challenging. Climate and weather conditions, dissected relief, soil and forest cover in this region impede acquisition of geological data. Mineral exploration requires development of fast and cost-effective methods that allow processing of insufficient and fragmented data like spaceborne images. Consideration of specific defoliation technique of directed principal components (DPC) and mineral mapping as well as application of assessment technique of Maximum Entropy (MaxEnt, also Logistic Regression, logit) led us to use Landsat 8 OLI for modeling of mineralization distribution in the Kunashir and Iturup Islands. DPC analysis is an image enhancement technique that processes two band ratio images. One of these images contains information of the component of interest (e.g., hydrothermal alteration), the other image contains information related to the spectrally interfering component (e.g., vegetation). Computed DPCs with loadings of opposite signs on either of input images, highlight unique contributions of each band ratio. Therefore, DPC analysis allows to decrease influence of noise related to vegetation. MaxEnt is a general-purpose technique of spatial prognosis based on incomplete data; its key feature is to assess the degree of target presence or absence by approximating the probability distribution of maximum entropy or nearest to uniform distribution based on a set of constraints taken from the incomplete data. Mapping based on DPC analysis in this work aimed to outline hydrothermal alteration associated with known mineralization. Division of the study area into zones according to their mineral prospectivity was then achieved with configured, trained and validated MaxEnt model using the datasets pertaining to a part of the Kunashir Island, which is better studied and explored. The trained predictive model was then applied to the less explored Iturup Island. As a result, the predictive model rendered a mineral prospectivity map, which correctly outlines the locations of the known mineralization. The application of the specific computation techniques to spectral data processing allows assessment of the contribution of Landsat 8 band ratios into the directed principal components and their implementation for cost effective prospectivity mapping including known and probable locations of unknown Au–Ag-bearing zones.

Mapping of Hydrothermal Alteration Zones in the Kelâat M’Gouna Region Using Airborne Gamma-Ray Spectrometry and Remote Sensing Data: Mining Implications (Eastern Anti-Atlas, Morocco)

The mapping of hydrothermal alteration zones associated with mineralization is of paramount importance in searching for metal deposits. For this purpose, targeting alteration zones by analyzing airborne geophysical and satellite imagery provides accurate and reliable results. In the Kelâat M’Gouna inlier, located in the Saghro Massif of the Moroccan Anti Atlas, natural gamma-ray spectrometry and ASTER satellite data were used to map hydrothermal alteration zones. Natural gamma-ray spectrometry data were processed to produce maps of Potassium (K in %), Uranium (eU in ppm), Thorium (eTh in ppm) and ratios of K/eTh and K/eU. In addition, four-band ratios were computed, on ASTER data, to map the distribution of clay minerals, phyllitic minerals, propylitic minerals, and iron oxides. The combined results obtained from geophysical and satellite data were further exploited by fuzzy logic modelling using the Geographic Information System (GIS) to generate a mineral prospectivity map. Seven hydrothermal alteration zones likely to be favorable for mineralization have been identified. They show a spatial correlation with (i) known surface prospects and mineral occurrences, (ii) the granite-encasing contact zone, and (iii) the fault zones (Sidi Flah and Tagmout faults). This research therefore provides important information on the prospecting of mineral potential in the study area.

Determination of Predictive Variables in Mineral Prospectivity Mapping Using Supervised and Unsupervised Methods

Determination of Predictive Variables in Mineral Prospectivity Mapping Using Supervised and Unsupervised Methods

Mapping prospectivity for regolith-hosted REE deposits via convolutional neural network with generative adversarial network augmented data

The regolith-hosted rare earth elements (REE) deposits are the dominant source of the global heavy REE resources. This study proposed a convolutional neural network (CNN) architecture to integrate the multi-source data (e.g., geological, geochemical and geomorphological data) to map mineral prospectivity of regolith-hosted REE deposits. To solve the lack of labelled data for the training of the supervised CNN, a generative adversarial network (GAN) was applied for data augmentation. The proposed GAN is trained in an unsupervised way, which regards the random downscaled real samples as the inputs. The GAN-based augmented data is validated by comparison with real data and the peak signal-to-noise ratio value, respectively. A case study of mapping prospectivity for regolith-hosted rare earth elements deposits in Southern Jiangxi Province of China further illustrates and validates the procedure. The final mineral prospectivity map is obtained by the CNN with all geological, geochemical and geomorphological data augmented by GAN. The CNN reaches 99.7% training accuracy and 98.9% validation accuracy. All the known mineral deposits are located in the prospective areas delineated by the CNN, which only occupy 2.36% of the study area. The obtained results indicate that the proposed framework, integrating CNN and the GAN-based augmentation method is an effective way for the application of the supervised deep learning algorithms in mineral prospectivity mapping of the regolith-hosted REE deposits and the prospective areas could be used for guiding further exploration in the study area.