A true positive is a positively labeled data point that a classification model correctly assigns to the positive class. The positive class depends on the problem: sick patients may be positive in a medical model, while spam emails may be positive in an email filter. section 7.5
Two everyday examples
A sick patient correctly diagnosed as sick is a true positive. Likewise, a spam email correctly classified as spam is a true positive. On a diagram, it is a positively labeled point that falls inside the model’s positive zone. section 7.5
Table 7.3 The confusion matrix of our second coronavirus model
|
Coronavirus model 2
|
Diagnosed sick
|
Diagnosed healthy
|
|
Sick
|
8 (true positives)
|
2 (false negatives)
|
|
Healthy
|
48 (false positives)
|
942 (true negatives)
|
A confusion matrix makes the count concrete: coronavirus model 2 has 8 true positives, meaning eight sick patients were correctly diagnosed. section 7.7
True positives are counted when measuring recall. Recall divides the number of true positives by all positively labeled cases, including both true positives and false negatives. section 7.7

