liveBook · Manning

False positive

As explained in Inside AI

A false positive is an alert or classification indicating a condition that is not actually present.

A fraud filter must catch criminal payments without wrongly rejecting genuine ones. A false positive is an alert or classification that identifies a condition—such as fraud—that is not actually present. section 11.11

In online payments, a false positive can mean turning away a legitimate customer. One in five rejected transactions was reported to be a false positive, and more than half of those customers later took their business elsewhere. section 11.11

False alarms also burden anti-money-laundering investigations: analysts may spend substantial time reviewing transactions flagged by fixed rules. Reducing false positives matters alongside detecting real fraud; neither goal should be considered alone. section 3.6 section 11.11

How Inside AI covers it

How deep it goes
Defines false positives and explains their commercial and investigative costs, without detailing how to measure or reduce them.
Examples
Uses wrongly rejected online purchases and rule-generated anti-money-laundering alerts as examples.
What you can do after
Understand why reliable fraud detection must limit false positives as well as catch actual fraud.
How it connects
Connects false positives to fraud detection, money laundering, supervised learning, unsupervised learning, and clustering.

Also explained in

Compare how 26 books explain false positive →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Inside AI. The section links point to the parts of the book it draws on.