ROC

Receiver Operating Characteristics (ROC) compares predicted and actual target values in a binary Classification model.

To edit and view an ROC:

1. Select the Target value. The ROC curves for that value are displayed.

2. Click Edit Custom Operating Point to change the operating point. The ROC graph displays a line showing ROC for each model. Points are marked on the graph indicating the values shown in the key at the bottom of the graph. Below the graph, the ROC Summary results table supplements the information presented in the graph. You can minimize the table using the splitter line.

   • The Models grid, in the lower pane, contains the following summary information:

     • Name

     • Area under the curve

     • Maximum Overall Accuracy Percentage

     • Maximum Average Accuracy Percentage

     • Custom Accuracy Percentage

     • Model Accuracy Percentage

     • Algorithm

     • Creation Date and Time

   • Select a model and click to see the ROC Detail Dialog box, which displays the statistics for probability thresholds.

   • To change the list of models, click to open the Edit Test Result Selection dialog box. By default, results for all models in the node are displayed.