The classification error Ei of an individual program
i depends on the number of samples incorrectly classified
(false positives plus false
negatives) and is evaluated by the formula:
where f is the number of sample cases incorrectly classified, and
n is the total number of sample cases.
|