DETERMINATION OF TECHNOLOGICAL OXIDATION ZONES AT URANIUM DEPOSITS IN KAZAKHSTAN USING MACHINE LEARNING METHODS MACHINE LEARNING

Abstract

The determination of technological acidification zones in uranium deposits during leaching is necessary for precise control and optimization of the uranium extraction process. Incorrect determination of the technological acidification zone can lead to excessive use of acidic reagents, which not only increases costs, but also can cause undesirable environmental consequences. The paper proposes an approach to solving issues related to the manual determination of zones of technological acidification in uranium deposits in Kazakhstan. The approach includes the study of machine learning algorithms to automate the identification of these critical areas. The use of artificial neural network (ANN) models and the extreme gradient boosting (XGB) model has shown its effectiveness in automating and improving the identification of these important zones during the mining of uranium deposits by underground borehole leaching. Thus, the accuracy of acidification intervals according to the F1-score metric for the ANN model is 0,75, and for the XGB model it is 0,80.