Scatterplot of response fits vs actual values

Note

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

Use the scatterplot to assess the accuracy of the predictions. When the analysis uses a validation technique, you can also compare the accuracy of the model for the training data and the validation results.

The scatterplot shows the actual response values on the x-axis and the fitted response values on the y-axis. The calculated line represents when the actual value and the response value are equal. Outliers or unusual clusters of points in the validation results can indicate data that deserve more investigation.

Interpretation

Ideally, the line goes through the middle of the points and the points are close to the line, relative to the scale of the response variable. When the analysis uses a validation method, the results include separate plots for the training data and for the validation results. The performance of the tree from the validation results is a typically a better representation of how the tree performs for new data. You should investigate large differences between the validation results and the training data.

In this scatterplot, the points for the training data and the cross-validation results sets show similar patterns. This similarity suggests that the performance of the model on new data is close to the performance of the model on the training data.

The scatterplot of fitted loan amounts versus actual loan amounts shows the relationship between the fitted and actual values for both the training data and the cross-validation results. You can hover over the points on the graph to see the plotted values more easily. In this example, most points fall approximately near the reference line of y=x.