Evaluation Step - Evaluation Tab

During the Evaluation step, use the options on the Evaluation tab to evaluate the models and make your final selection.

Define cost matrix. Click this button to identify the cost of misclassifying an observation.

Evaluate models. Click this button to generate a results workbook that contains various summaries of deployment results for the selected models. The results include error rates for each model, predictions and residuals, correlations between each model and the selected variable, and a scatterplot.

Select model(s). Use the options in this grid to review model statistics and select the model you want to use. To select a model, select the check box adjacent to the ID number in the ID column.

ID. All DMR models have a unique identity number. This will allow you to quickly recognize and uniquely identify the individual models in the latter stages of the DMR analysis.

Name. Each model has a name depending on its type. These include C&RT, Random Forest, Neural Network, Boosting Trees, and SVM.

Error rate. Percentage of cases that are incorrectly predicted by a model.