Definition False positive

Machine learning algorithms can be used to classify an object into two different categories. For instance, spam filters can be used to scan e-mails and identify them as legitimate ("not spam") or junk ("spam”). In this context, a positive classification is the correct identification and sorting of an unsolicited e-mail item into the “spam” category. A false positive is the misidentification and sorting of a legitimate e-mail item into the unsolicited (“spam”) e-mail category.