In statistical classification, the accuracy of a classifier is the ratio between the number of correct predictions and the number of predictions.

$\text{accuracy} = \frac{N_\text{correctly predicted}}{N}$

See also Recall, Precision and F-score.