Given that dilution is a problem that reduces the profit of many underground operations, the dilution monitoring, prediction, and reduction convey the capability to increase operational efficiency. The proposed research aims to develop an unplanned dilution monitoring, prediction, and reduction tool. The research methodology will be based on artificial intelligence (AI)-based models. More specifically, the research will focus on the applicability of ensemble classifiers, naive Bayes classifiers, logistic regression, and neural networks.