A New Method to Evaluate Gold Mineralisation-Potential Mapping Using Deep Learning and an Explainable Artificial Intelligence (XAI) Model

Abstract

Geoscientists have extensively used machine learning for geological mapping and exploring the mineral prospect of a province. However, the interpretation of results becomes challenging due to the complexity of machine learning models. This study uses a convolutional neural network (CNN) and Shapley additive explanation (SHAP) to estimate potential locations for gold mineralisation in Rengali Province, a tectonised mosaic of volcano-sedimentary sequences juxtaposed at the interface of the Archaean cratonic segment in the north and the Proterozoic granulite provinces of the Eastern Ghats Belt in Eastern India. The objective is to integrate multi-thematic data involving geological, geophysical, mineralogical and geochemical surveys on a 1:50 K scale with the aim of prognosticating gold mineralisation. The available data utilised during the integration include aero-geophysical (aeromagnetic and aerospectrometric), geochemical (national geochemical mapping), ground geophysical (gravity), satellite gravity, remote sensing (multispectral) and National Geomorphology and Lineament Project structural lineament maps obtained from the Geological Survey of India Database. The CNN model has an overall accuracy of 90%. The SHAP values demonstrate that the major contributing factors are, in sequential order, antimony, clay, lead, arsenic content and a magnetic anomaly in CNN modelling. Geochemical pathfinders, including geophysical factors, have high importance, followed by the shear zones in mineralisation mapping. According to the results, the central parts of the study area, including the river valley, have higher gold prospects than the surrounding areas. Gold mineralisation is possibly associated with intermediate metavolcanics along the shear zone, which is later intruded by quartz veins in the northern part of the Rengali Province. This work intends to model known occurrences with respect to multiple themes so that the results can be replicated in surrounding areas.

A Country Wide Evaluation of Sweden's Spatial Flood Modeling With Optimized Convolutional Neural Network Algorithms

“…Pradhan et al. (2023) used CNN models to detect flood‐prone areas in Jinju Province, South Korea. The results showed acceptable predictive performance of model with the accuracy of 88.4%.…”

A comparative spatial analysis of flood susceptibility mapping using boosting machine learning algorithms in Rathnapura, Sri Lanka

“…The explainability of ML models comes under explainable artificial intelligence (XAI) which demonstrates the importance of conditioning factors to overall predicted results, allowing the analyst to determine what the ML model considers when assessing flood susceptibility (Aydin & Iban, 2022; Pradhan et al, 2023). In this study, a local XAI approach namely shapley additive explanation (SHAP) technique is employed to gain insights into the primary factors that influence predictions and their impact on individual predictions (sample‐wise scale).…”

“…In this research, statistical indices such as F1‐score, kappa coefficient and area under curve (AUC), were used to evaluate the model performances, which are widely used in geohazard modeling studies (Pradhan et al, 2023; Seleem et al, 2022; Zhao et al, 2021). The F1‐score is a statistical measure used in binary classification to evaluate the accuracy of a model's predictions and calculated as the harmonic mean of precision and recall, providing a balance between these two metrics.…”

“…These models have many hyperparameters and optimizing them can be challenging and time‐intensive task (Seydi et al, 2023). Although data‐driven models generate more accurate outcomes, they are considered to be “black boxes” due to limited interpretability, making them unpopular with stakeholders (Pradhan et al, 2023). In previous studies, data‐driven models for FSM were commonly validated within the same geographical region employed for training the models.…”

Automatic Explanation Of Metal’s Phenomena By Application Of Artificial Neural Networks – Use Case For Hot Extruded Aluminum Alloy 6082

“…It can be concluded that it represents a substantial advancement and simplification based on the prior studies [1,3,25,33], where the explanations/conclusions were typically produced on the basis of an expert assessment of the obtained results. Compared to existing XAI approaches [34][35][36], the suggested framework attempts to imitate the standard scientific methodology, which is based on mathematical representations of the considered physical phenomena with the lowest number of parameters and its derivatives. Furthermore, using the predicted values, we attempt to analyze the reliability and accuracy of the results, which are related to well-known statistical metrics.…”