Specify the class weights for Fit Model and Discover Key Predictors with TreeNet^® Classification

Predictive Analytics Module > TreeNet® Classification > Fit Model > Class weights

Predictive Analytics Module > TreeNet® Classification > Discover Key Predictors > Class weights

Note

This command is available with the Predictive Analytics Module. Click here for more information about how to activate the module.

Class weights allow you to specify the weights for each class of the response variable. For example, an observation from one class may weigh more than an observation from a different class. Class weights are similar to prior probabilities and can help provide a more accurate model.

Note

Class weights are different from individual case weights. You can use both types of weights in the same analysis. For more information on individual case weights, go to Select the analysis options for Fit Model and Discover Key Predictors with TreeNet® Classification.

Unweighted: Minitab does not weigh the classes.
Weighted to ensure equal sample sizes across classes: Minitab automatically balances unequal class sizes so that the class weights are proportional to the sample frequencies of classes.
Specify weights: Enter a weight for each class. Class weights affect how the model is generated and the performance of the model, but does not change the class counts.
Class weights must be > 0. The default value is 1.

Specify the class weights for Fit Model and Discover Key Predictors with TreeNet® Classification

Note

Note

Specify the class weights for Fit Model and Discover Key Predictors with TreeNet^® Classification