Deep Learning · Level 2 of 5
Batch Normalization
Normalizing layer activations using batch statistics with learned scale and shift.
Training usually uses batch statistics while evaluation uses accumulated estimates.
Example
A vision layer standardizes each channel before applying its learned scale.
Listen to the definition and example
Audio transcript
Batch Normalization. Normalizing layer activations using batch statistics with learned scale and shift. Training usually uses batch statistics while evaluation uses accumulated estimates. For example: A vision layer standardizes each channel before applying its learned scale.
Explore this concept
Why it matters
This helps explain why a neural model learns well, becomes unstable, or fails to generalize.
Start with
Quick recall question
Try answering before looking back at the definition.