Misclassification table for CART^® Classification

In This Topic

Input Misclassification Cost
Actual Class
Misclassed Training
Misclassed Test

Input Misclassification Cost

Indicates the misclassification costs that you specify. By default, Minitab uses equal costs of 1. To indicate higher costs, use larger values. Costs must be greater than 0.

Interpretation

For example, suppose the cost of misclassifying a potential customer is 10 times more costly than misclassifying an uninterested customer.

Predicted Level
Actual Level	Yes (Event)	No
Yes (Event)		10
No	1

Actual Class

Indicates the counts of the actual events.

Misclassed Training

Misclassed: Indicates the number of cases that were misclassified in the training data set.
%Error: Indicates the percentage of cases that were misclassified in the training data set.

Cost: Indicates the cost of the cases that were misclassified in the training data set. The cost is the product of the %error * misclassification cost.

Misclassed Test

Misclassed: Indicates the number of cases that were misclassified in the test data set.
%Error: Indicates the percentage of cases that were misclassified in the test data set.

Cost: Indicates the cost of the cases that were misclassified in the test data set. The cost is the product of the %error * misclassification cost.