A statistical procedure to assess how well a model, fit using a training dataset, predicts outcomes in a holdout dataset, drawn from the same population as the training dataset.
« Back to Glossary Index « Back to Glossary Index
Automated Valuation Metrics