Patent · US Active

Template regularization for generalization of learning systems

US9390382B2 · kind B2 · utility

2Cited by

0References

20Claims

0Family size

Assignee

Google LLC · US

Inventors

Yoram Singer · Palo Alto, US
Tal Shaked · Mountain View, US
Tushar Deepak Chandra · Los Altos, US
Tze Way Eugene Ie · San Francisco, US
James Vincent McFadden · San Francisco, US
Jeremiah Harmsen · San Jose, US
Kristen Lefevre · Ann Arbor, US

Key dates

Filing date	Dec 30, 2013
Grant date	Jul 12, 2016
Priority date	—
Expiry date	Jun 27, 2034

Classification

Technology area (CPC G)Physics
CPC primaryG06N20/00
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

Systems and techniques are disclosed for training a machine learning model based on one or more regularization penalties associated with one or more features. A template having a lower regularization penalty may be given preference over a template having a higher regularization penalty. A regularization penalty may be determined based on domain knowledge. A restrictive regularization penalty may be assigned to a template based on determining that a template occurrence is below a stability threshold and may be modified if the template occurrence meets or exceeds the stability threshold.

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