Prediction Confidence
Robustness indicator of a classification or regression predictive model.
The prediction confidence indicates the capacity of the predictive model to achieve the same performance when it is applied to a new data source, which has the same characteristics as the training data source. If the distribution of data is different between the two data sources, the predictive model is no longer useful.
The prediction confidence indicator takes a value between 0% and 100%.
A predictive model with a prediction confidence that is equal to or greater than 95% is considered robust. It has a high capacity for generalization.
A prediction confidence less than 95% should be considered with caution. If you apply the predictive model to a new data source, you risk generating unreliable results.
The prediction confidence indicator is color-coded: A value greater than or equal to 95% appears in green, while a value less than 95% appears in red.