Patent · US Active

Training neural networks using posterior sharpening

US10824946B2 · kind B2 · utility

3Cited by
1References
20Claims
0Family size

Assignee

Inventors

Key dates

Filing dateJul 15, 2019
Grant dateNov 3, 2020
Priority date
Expiry dateJul 15, 2039

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06N3/09
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a neural network. In one aspect, a method includes maintaining data specifying, for each of the network parameters, current values of a respective set of distribution parameters that define a posterior distribution over possible values for the network parameter. A respective current training value for each of the network parameters is determined from a respective temporary gradient value for the network parameter. The current values of the respective sets of distribution parameters for the network parameters are updated in accordance with the respective current training values for the network parameters. The trained values of the network parameters are determined based on the updated current values of the respective sets of distribution parameters.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.