A computer-based method for creating a personalized neural network model.

Abstract

This patent describes a method by which a machine learns to create a personalized neural network model, involving the step of receiving one or more personal data samples from a user. A prototype of the personal identity is created based on these personal data samples. The prototype of the personal identity is trained to reflect the user's personal characteristics. A network graph is created based on the prototype of the personal identity. One or more channels of the general network are truncated based on the network graph to create a personalized model. Additionally, the term "truncated" refers to the machine used to create the personalized network model.