We can simply measure the number of correct decisions our classifier makes, divide by the total number of test examples, and the result is the accuracy of your classifier. It’s that simple. The vast majority of research results report accuracy, and many practical projects do too.