The ground reference dataset used in the assessment of classification accuracy is typically
assumed implicitly to be perfect (ie, 100% correct and representing ground truth). Rarely is …