Assessment of the mineral potential of rocks using intelligent data analysis technologies

This paper presents a study aimed at developing and evaluating a classification model for the mineral potential of geological layers using machine learning methods. The main goal was to create an algorithm capable of predicting the categories of mineral potential — low, medium and high — based on key geological characteristics such as ore content, density, depth, permeability and rock type. The results obtained demonstrated the overall accuracy of the model at 98%, with an almost ideal classification for the categories of medium and high mineral potential. The analysis of the significance of the features showed that the ore content and density have the greatest impact on the classification. The work also revealed certain difficulties in classifying low mineral potential, which is due to limited data or the intersection of characteristics between classes. These results highlight the importance of further research in feature selection and improving data balance. The findings demonstrate the effectiveness of using machine learning methods in geological research and their potential for automating mineral resource assessment processes. The developed methodology can be adapted to solve other problems in the field of geology, such as forecasting deposits and optimizing mining.