---

A Hybrid Framework for Detecting Gold Mineralization Zones in GR Halli, Western Dharwar Craton, Karnataka, India.

Raju PVS, Mudili VS, Ganivada A.

MINERALS.

https://doi.org/10.3390/min15111125 

AI and machine learning techniques were used to map areas with high potential for valuable mineral deposits. This study introduces a hybrid model integrating WGAN-GP for generating balanced synthetic negative samples, CNN for feature extraction, and FKELM for classification, tackling imbalanced and uncertain data in mineral exploration. Applied to the G.R. Halli gold prospect in India's Chitradurga Schist Belt using nine geochemical elements, it outperformed SVM, Gradient Boosting, and baseline CNN. FKELM achieved AUC 0.976 and 92% accuracy, yielding coherent maps aligned with known gold zones and anomalies, offering a scalable solution for complex terrains.

Showing the Mineral prospectivity mapping using Fuzzy Kernel Extreme Learning Machines.

---