Machine Learning aplicado a la exploración minera usando matriz de confusión

The objective of this study was to determine how to spatially validate geological datasets obtained from the Huancayo region in Peru, as well as to identify the most common methods for evaluating the performance of classification models. The methodology employed involved the use of machine learning algorithms applied to a confusion matrix, with results represented through a ROC curve. The findings indicated that the accuracy parameter reached a maximum percentage of 65.4%, precision was 80.4%, recall was 96.2%, and the F1-score achieved a value of 96.1%. It is concluded that when using classification models, it is essential to consider additional metrics beyond accuracy to obtain a more reliable estimation.