Some statistical methods for inference have been developed for the F 1 score in binary classification problems; however, they have not been extended to the problem of multi-class classification. There are three types of F 1 scores, and statistical properties of these F 1 scores have hardly ...
imblearn.metrics.macro_averaged_mean_absolute_error() raises an exception when not all classes are represented in the ground truth. Thought I'd flag this because a similar function in sklearn behaves differently to imbalanced-learn. That function is f1_score with macro averaging. Here's an exam...
F-Score (Balanced ) as the harmonic mean of the Precision and Recall : Micro Averaged metrics given two different set of data : Macro Averaged metrics with two datasets : Macro-averaged metrics are used when we want to evaluate systems performance across on different datasets. Micro-averaged me...
Micro-averaged result (excluding Other): P = 2055/2311 = 88.92% R = 2055/2263 = 90.81% F1 = 89.86% MACRO-averaged result (excluding Other): P = 88.64% R = 90.27% F1 = 89.43% <<< The official score is (9+1)-way evaluation with directionality taken into account: macro-averaged F1...