Models will be scored based on accuracy (percentage of classifications that are correct), which is one minus the misclassification rate (percentage of classifications that are wrong).