“…Based on these items further indicators are to be defined, as PA is the producer's accuracy (recall or true positive rate or sensitivity), UA the user's accuracy (consumer's accuracy or precision or positive predictive value), OA the (overall) accuracy, OE the omission error (false negative rate), CE the commission error (false discovery rate), TNR the true negative rate, FPR the false-positive rate, NPV the negative predictive value and finally FOR the false omission rate. The rate values are often given in per cent (Barsi et al 2018). Confusion matrix can be extended also for multiclass cases.…”