	US 12,265,888 B1
	Edge computing for machine learning
Gregory Brian Meyer, San Antonio, TX (US); Mark Anthony Lopez, Helotes, TX (US); Ravi Durairaj, San Antonio, TX (US); Nolan Serrao, Plano, TX (US); Victor Kwak, Frisco, TX (US); Ryan Thomas Russell, San Antonio, TX (US); Christopher Russell, The Colony, TX (US); and Ruthie D. Lyle, Durham, NC (US)
Assigned to United Services Automobile Association (USAA), San Antonio, TX (US)
Filed by UIPCO, LLC, San Antonio, TX (US)
Filed on May 28, 2020, as Appl. No. 16/885,327.
Claims priority of provisional application 62/855,095, filed on May 31, 2019.
Int. Cl. G06N 20/00 (2019.01)

CPC G06N 20/00 (2019.01)

20 Claims

1. A method for performing machine learning in mobile computing devices which are configured to be coupled with a cloud computing system, the method comprising:

receiving at each one of the mobile computing devices a local machine learning system from a plurality of local machine learning systems in a global machine learning system hosted by the cloud computing system, the plurality of local machine learning systems based on different machine learning models and each local machine learning system is received by a different mobile computing device;

activating, on each mobile computing device, a machine learning application, which accesses the local machine learning system including a local machine learning model, which includes one or more training parameters, and which is stored in a memory of each mobile computing device;

periodically updating each local machine learning system based upon updates for the local machine learning system received from the global machine learning system hosted by the cloud computing system;

performing machine learning on the local machine learning system at each mobile computing device at the same time, using training data received at each mobile computing device; and

periodically transmitting changes to the local machine learning system resulting from the machine learning performed at each mobile computing device from each mobile computing device to the global machine learning system hosted by the cloud computing system to update the local machine learning system in the plurality of local machine learning systems in the global machine learning system;

wherein the local machine learning system at one of the mobile computing devices is received from a different version of the global machine learning system, and each local learning model is stored and updated at a given mobile computing device and the cloud computing system.