Relative prediction error of flotation indices by anfis models application/pdf 1562004 bytes All rights reserved Jovanović, Ivana Conić, Vesna Milanović, Dragan Nakhaei, Fardis Krstić, Slađana 2023 info:eu-repo/semantics/conferenceProceedings Keywords: flotation modeling, relative prediction error, ANFIS. eng Proceedings International Mineral Processing and Recycling Conference (15 ; 2023 ; Belgrade) startpage: 387 endpage: 390 Abstract: This paper describes performances of three different ANFIS models trough the important statistical parameters – relative prediction error, its standard deviation and its mean. Copper content in the ore feed, collector dosage and frother dosage were selected as input parameters to estimate the copper grade and recovery in the final concentrate, and the copper content in the final tailings of the flotation plant. For each output parameter one ANFIS model is developed. Evaluation of the proposed models were performed on the basis of real process data collected by the multiannual monitoring of industrial flotation plant operation in “Veliki Krivelj Mine”. The results showed that the proposed ANFIS methodology is the most suitable for prediction of copper recovery in the final concentrate, given that the smallest relative prediction errors with their smallest standard deviation (SDE = 0.11) were obtained in this case. https://unilib.phaidrabg.rs/o:5973 cobiss:138181897 ISBN: 978-86-6305-133-1