July 2019
Beginner to intermediate
740 pages
16h 52m
English
Using the values in the confusion matrix, we can calculate metrics to help evaluate the performance of a classifier. The best metrics will depend on the goal for which we are building the model and whether our classes are balanced. The formulas in this section are derived from the data we get from the confusion matrix, where TP is the number of true positives, TN is the number of true negatives, and so on.
Read now
Unlock full access