English (English)

Classification quality metrics

The following metrics can be used to assess the quality of classification:

  • Precision – This metric is an assessment of the classifier's ability to assign classes precisely to those documents that actually belong to those classes. However, this metric does not account for overlooked documents.
  • Recall – This metric is an assessment of the classifier's ability to assign classes to documents without overlooking relevant documents. However, this metric does not account for misclassified documents.
  • F-measure – This metric is a combination of precision and recall.
  • Confidence score – This metric is an assessment of the probability that a document belongs to a particular class.

25.05.2023 7:55:03

Please leave your feedback about this article

Usage of Cookies. In order to optimize the website functionality and improve your online experience ABBYY uses cookies. You agree to the usage of cookies when you continue using this site. Further details can be found in our Privacy Notice.