“System and method for addressing overfitting in a neural network”

​

2016-08-02

Application granted

2017-10-05

Assigned to GOOGLE LLC

2019-06-25

Application status is Active

A system for training a neural network. A switch is linked to feature detectors in at least some of the layers of the neural network. For each training case, the switch randomly selectively disables each of the feature detectors in accordance with a preconfigured probability. The weights from each training case are then normalized for applying the neural network to test data.

https://patents.google.com/patent/US9406017B2/en