Glossary term
Glossary term
Foundations
Normalizing the input or output of the activation functions in a hidden layer. Batch normalization can provide the following benefits:
Make neural networks more stable by protecting against outlier weights.
Enable higher learning rates, which can speed training.
Reduce overfitting.
Created for this library
A computer vision team adds batch normalization to its CNN to allow higher learning rates without diverging during training.
A speech recognition vendor uses batch normalization in its acoustic model and reports faster convergence and better generalization.
A medical imaging team adopts batch normalization in its segmentation network because it stabilizes training across heterogeneous scanner data.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License