# classificationError The fraction of incorrectly labeled examples. ``` var classificationError: Double { get } ``` ## Discussion The classification error describes how many examples were incorrectly labeled divided by the total number of examples. Accuracy as a percentage may be more intuitive. You can calculate it as follows: ```swift let accuracy = (1 - metrics.classificationError) * 100 ``` > Important: This is a useful metric only when the data is well-balanced between categories. For example, > suppose you build a classifier to detect a rare disease with very few examples of sick patients compared > to the number of healthy patients. Predicting that a new patient will always be healthy would be highly > accurate (low classification error), but a poor disease detector. The ``doc://com.apple.createml/documentation/CreateML/MLClassifierMetrics/precisionRecall`` > and ``doc://com.apple.createml/documentation/CreateML/MLClassifierMetrics/confusion`` properties provide more detail in these cases. --- Copyright © 2026 Apple Inc. All rights reserved. | [Terms of Use](https://www.apple.com/legal/internet-services/terms/site.html) | [Privacy Policy](https://www.apple.com/privacy/privacy-policy)