Accuracy: given a classification algorithm and a test set, the accuracy is the proportion of correct predictions done by the algorithm over all predictions in the test set