Exploration
Feridon Ghadimi; Arman Ghadimi
Abstract
The Geochemical Mineralization Probability Index (GMPI) represents a phase weight assigned to each geochemical stream sediment sample for individual indicator components. In this framework, the weights of the evidence map classes are determined based on the factor scores (FS) derived from factor analysis ...
Read More
The Geochemical Mineralization Probability Index (GMPI) represents a phase weight assigned to each geochemical stream sediment sample for individual indicator components. In this framework, the weights of the evidence map classes are determined based on the factor scores (FS) derived from factor analysis for each indicator component. A wide range of direct and indirect approaches has been proposed to delineate prospective zones for mineral exploration; however, many of these methods are associated with substantial time and financial costs. Accordingly, this study aims to evaluate the performance of an Artificial Neural Network (ANN) model and to compare its predictive capability with that of conventional model-architecture-based approaches. In this research, a hybrid Artificial Neural Network–Biogeography-Based Optimization (ANN–BBO) model is developed to estimate the GMPI. The dataset used for model training and validation comprises geochemical concentrations of Au, Cu, Pb, Zn, Ag, Mo, W, and Sn obtained from 109 stream sediment samples collected in the Zaghar area. Biogeography-Based Optimization (BBO) is employed to optimize the ANN training process by adaptively adjusting its parameters to enhance predictive performance. The proposed ANN–BBO model achieved a Mean Squared Error (MSE) of 0.0221 and a coefficient of determination (R²) of 0.8244, indicating satisfactory predictive accuracy. Furthermore, the model demonstrated robust generalization capability, maintaining reliable predictive performance despite the considerable geochemical variability observed within the stream sediment dataset. The results of the sensitivity analysis reveal that Pb and Ag exert the most significant influence on the prediction of geochemical anomalies within the study area.
