I have been trying to understand applying the back propagation on batch-normalization. However, the matrix multiplication and equations involved have got me confused and I feel that my understanding of the forward pass itself is mistaken.

To highlight my understanding on the batch normalization forward pass and backward pass, I have written the equations in an IPython Notebook available here.

In the notebook I have explained the corresponding matrix multiplications and output-dimension and the point where mu understanding is not aligned with that explained in the paper.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.