Confusion matrix for CART^® Classification

Find definitions and interpretations for every statistic in the Confusion matrix.

The Confusion matrix shows how well the tree separates the classes correctly using these metrics:

True positive rate (TPR) — the probability that an event case is predicted correctly
False positive rate (FPR) — the probability that a nonevent case is predicted incorrectly
False negative rate (FNR) — the probability that an event case is predicted incorrectly
True negative rate (TNR) — the probability that a nonevent case is predicted correctly

Interpretation

		Predicted Class (Training)			Predicted Class (Cross-validation)
		Predicted Class (Training)			Predicted Class (Cross-validation)
Actual Class	Count	Yes	No	% Correct	Yes	No	% Correct
Yes (Event)	139	117	22	84.2	105	34	75.5
No	164	22	142	86.6	24	140	85.4
All	303	139	164	85.5	129	174	80.9

Statistics	Training (%)	Cross-validation (%)
True positive rate (sensitivity or power)	84.2	75.5
False positive rate (type I error)	13.4	14.6
False negative rate (type II error)	15.8	24.5
True negative rate (specificity)	86.6	85.4

In this example, the total number of Yes events is 139, and the total number of No events is 164.

In the training data, the number of predicted Yes events is 117, which is 84.2% correct.
In the training data, the number of predicted No events is 142, which is 86.6% correct.
In the cross-validation results, the number of predicted Yes events is 105, which is 75.5% correct.
In the cross-validation results, the number of predicted No events is 140, which is 80.9% correct.

Overall, the %Correct for Training is 85.5% and 80.9% for Cross-validation.