چكيده فارسي :
Iron ore concentrates due to their main impacts on the process of iron making have been mainly attended in steel industries. Accordingly, their contained species as well as the structural properties especially the fine size of them directly influence on the properties of the products obtained from iron making process. Hence, the origin of iron ore should be controlled. To control the properties of feed and prevent the cheating, five various iron ore concentrates, provided from different mines including Bafgh, Golgohar, Goharzamin, Kimia, and Chadormaloo were studied. So, the content of total Iron and Ferrous were determined by classical chemical method of oxidation-reduction titration. In addition, the size of them was measured by calibrated relevant sieves. Moreover, their analytical compositions were analyzed using X-ray fluorescence spectroscopy (XRF). Furthermore, the sulfur contents were also determined using combustion sulfur analyzer. The data was employed for pattern recognition using principal components analysis (PCA) as a clustering method and both partial least squares discriminant analysis (PLS-DA) and extended canonical variates analysis (ECVA) as classification methods. The dataset was split into the training and test sets to evaluate the classification performances using both cross-validation and external prediction. Classification was achieved successfully with more 90% accuracy.
