The accuracy tab in the middle panel similarly shows the accuracy measures of the selected model on the right panel.
The accuracy panel has four graphs;
True Positive: This shows the proportion of the target class predicted correctly. For instance, 79.9% Churn cases have been correctly predicted by the selected Model.
False Negative: This shows the proportion of target class predicted incorrectly as other classes. For instance, 20.2% of the churn cases have been incorrectly predicted as other classes by the model.
False Positive: This shows the proportion of other classes that is being incorrectly predicted as the target class. 32% of the non-churn customers have been incorrectly predicted as churns by the model.
True Negative: This shows the proportion of other classes that is being correctly predicted as other classes as it is supposed to. For instance, 67% of non-churn customers have been correctly predicted as the non-churn customers by the model.