You have trained a classification model with a highly sophisticated Machine Learning algorithm. Right. It is now time to evaluate its performance on test data, i.e. to score it.
A number of scoring metrics have been proposed over the years in different domains: sensitivity and specificity, precision and recall, accuracy, area under the curve, Cohen’s Kappa, and many more. Generally, they are based on values reported in a confusion matrix.
These slides are from a webinar we presented where we explore the concept of confusion matrix, true/false positives/negatives, and the related, most commonly used scoring metrics for classification models. We also demonstrate how to calculate all those metrics within KNIME Analytics Platform. https://www.knime.com/knime-software/knime-analytics-platform
View the webinar here: https://youtu.be/dOqRjeOv1VA