TRAINING A NEURAL NETWORK
A method (100) performed by a central processor of training a first neural network to classify medical data comprises receiving (102) a first parameter indicative of a weight in the first neural network, the first parameter being obtained during training of the first neural network on a first set of...
Saved in:
Main Author | |
---|---|
Format | Patent |
Language | English French German |
Published |
21.08.2019
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | A method (100) performed by a central processor of training a first neural network to classify medical data comprises receiving (102) a first parameter indicative of a weight in the first neural network, the first parameter being obtained during training of the first neural network on a first set of medical data. The method further comprises combining (104) the first parameter with a second parameter to determine an update to the weight, the second parameter being indicative of a respective weight in a second neural network. The method (100) further comprises sending (106) the update to the weight to the first neural network. |
---|---|
Bibliography: | Application Number: EP20180156991 |