Developing Machine Learning, Deep Learning, Clustering and Optimization Techniques for Intelligent Mineral Prospectivity Mapping and Geochemical Anomaly Modeling

Recently, machine learning algorithms (e. g., artificial neural networks (ANNs), support vector machines (SVM), random forest (RF), etc.), deep learning algorithms (e.g., convolutional neural network (CNN), recurrent neural network (RNN), long-short term memory (LSTM), etc.) and clustering algorithms (e. g., K-means, Fuzzy c-means (FCM), etc.) have been broadly applied into mineral prospectivity mapping (MPM) and geochemical anomaly modeling (GAM) for exploration of undiscovered deposits. However, the application of supervised machine learning and deep learning techniques as well as unsupervised clustering techniques may cause some challenging issues for MPM and GAM. For instance, inaccurate or incorrect selection of hyperparameters which are responsible to control the learning process in machine learning and deep learning algorithms may propagate systemic uncertainties in MPM and GAM. Therefore, several metaheuristic optimization algorithms (e. g., genetic algorithm (GA), harmony search (HS) algorithm, particle swarm optimization (PSO) algorithm, etc.) have been proposed to support machine learning, deep learning and clustering techniques in discovering faster and more accurate solutions.

The aim of this Research Topic is to focus on original works on various disciplines of intelligent MPM and GAM based on machine learning, deep learning, clustering and optimization techniques.

Potential topics include but are not limited to the following:
• Optimization of geochemical anomaly detection using a combination of novel clustering and metaheuristic optimization algorithms.
• Mapping of geochemical indicators based on incorporation of novel clustering and metaheuristic optimization algorithms into catchment basin analysis.
• Combination of popular geochemical anomaly detection approaches (e.g., concentration-area (C-A) fractal, singularity index (SI), spectrum-area (S-A) multifractal, etc.) and novel clustering and metaheuristic optimization algorithms for GAM.
• Application of various machine learning, deep learning and optimization algorithms for predictive modeling of mineral prospectivity as well as lithological mapping.
• Spatial predictive mapping of mineral prospectivity by automated tuning of hyperparameters related to supervised machine learning and deep learning algorithms using metaheuristic optimization algorithms.
• Spatial predictive mapping of mineral prospectivity by automated tuning of hyperparameters related to unsupervised clustering algorithms using metaheuristic optimization algorithms.
• Enhancing success of targeting for ore-related deposits by sensitivity analysis of prospectivity modeling to evidence maps.
• Quantification of systemic uncertainties linked to ore-related targets of supervised data-driven MPM.