A plethora of fraud detection (FD) systems based on big data are in the marketplace today. These systems require vast amounts of disjointly distributed data from multiple geolocations. However, FD is needed by many mobile applications. They require small memory footprints and fast computational times for efficient operation. Lightweight personalized analytics (LPA) is a technology designed to use individual analytics to gain insights for developing personalized offerings. It leverages a small number of analytics parameters to enable time and space efficiencies. It provides insights for enabling optimal FD results in mobile systems that require small data. The FD methods include Continuous User Authentication for Fraud Detection, and federated learning. FD methods, coupled with LPA (FD-LPA), may leverage as input user communication messages and usage behavior to detect anomalies. The technology can be used in a plethora of consumer-focused industries, some of which include military, finance, insurance, retail, healthcare, and automobile. GMF has filed 7 provisional patent applications and is drafting several more to build a patent portfolio on LPA technology.
