MAKİNE ÖĞRENME VE DERİN ÖĞRENME KULLANILARAK DEMİR CEVHERİNİN KÖPÜK FLOTASYON SÜRECİNDE KALİTE TAHMİNİ

Flotation is an enrichment method that separates materials by using the hydrophilic (water-loving) and hydrophobic (water-hating) properties of the ore to be produced, creating bubbles in a liquid and making the ore float or sink in water. In iron ore mining, stakeholders at the flotation plant rely on a traditional laboratory test technique, which usually takes more than two hours, to determine the two relevant variables needed to achieve the desired quality. Therefore, this study has used machine learning and deep learning techniques to real-time predict the percentage of silica concentrate (SiO2) in the flotation plant. The prediction model was created using the “Mining Process Flotation Plant Database” dataset obtained from Kaggle. A linear regression model and two different artificial neural network models were used, and the results were compared.