Accuracy is a measure of the percentage of all samples (a single piece of data is called a sample. For example, in customer data, it refers to customers) in the evaluation data that can be correctly identified.

For example, if “(a) Withdrawal” is set as the prediction value, and the following confusion matrix is obtained,

Out of the total number of samples (4 + 5 + 6 + 7 = 22), the number that correctly predicted “(a) Withdrawal;” “(b) Continuation” is 4 + 7 = 11. In this case, Accuracy is 50%, because 11/22 = 0.5.