liveBook · Manning

Credit card fraud

As explained in Outlier Detection in Python

Unauthorized use of a credit card, including use after loss, theft, or counterfeiting.

Credit card fraud is the unauthorized use of a card, commonly after loss, theft, or counterfeiting. Detection focuses on purchases or purchase sequences that differ from normal behavior, while recognizing that unusual activity is not automatically fraudulent. section 1.1.2

Outlier detection supports fraud screening by combining known behavioral rules with searches for unusual activity that predefined rules may miss. Tests can be general, cardholder-specific, merchant-specific, time-sensitive, or adjusted for special events. section 1.1.2

Illustrative fraud indicators include:

  • Two purchases occur far apart geographically within a short time window. section 1.1.2
  • An application uses a nonexistent address, invalid phone number, or nonsensical email address. section 1.1.2
  • A purchase conflicts with the cardholder’s history, merchant history, purchase time, or expected seasonal context. section 1.1.2

How Outlier Detection in Python covers it

How deep it goes
Describes general and contextual detection tests but stops short of specifying algorithms, implementation, or evaluation procedures.
Examples
Uses geographically distant purchases, invalid application details, unusual cardholder behavior, merchant history, and seasonal events.
What you can do after
Diagnose fraud risk by applying general rules and comparing activity with cardholder, merchant, temporal, and event-specific norms.
How it connects
Connects credit card fraud with outlier detection and financial fraud.

Related concepts

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