Course 1 Week 2 - ๐๐š๐œ๐ค๐ฉ๐ซ๐จ๐ฉ๐š๐ ๐š๐ญ๐ข๐จ๐ง ๐Ÿ๐จ๐ซ ๐‹๐จ๐ ๐ข๐ฌ๐ญ๐ข๐œ ๐‘๐ž๐ ๐ซ๐ž๐ฌ๐ฌ๐ข๐จ๐ง ๐–๐ข๐ญ๐ก๐จ๐ฎ๐ญ ๐ญ๐ก๐ž ๐Œ๐š๐ ๐ข๐œ

Hello everyone,

Iโ€™m currently taking the Neural Networks and Deep Learning course, and since Iโ€™m one of those people who canโ€™t fully trust that Iโ€™ve understood something until every detail โ€œclicks into placeโ€, I decided to attempt a stepโ€‘byโ€‘step derivation of the backpropagation formulae from scratch.

I created a supplement document that Iโ€™m sharing below for anyone who might be interested.

A few disclaimers before you dive in:

  • This derivation is heavy on vector calculus. Like, โ€œI had to blow dust off old neuronsโ€ heavy.
  • Following the notations in the course led me to some transposed Jacobians, andโ€ฆ those are not very pretty.
  • My vector calculus skills were a bit rusty going in, so this may not be the most elegant derivation ever written, but everything seemingly checks out in the end.
  • No LLMs were used in producing the document. So if you find mistakes, those are 100% my own. Please flag them so I can improve.

PS: I canโ€™t say Iโ€™m proud of my handwriting on a tablet, but I am proud of the document itself. I havenโ€™t yet seen a derivation this detailed anywhere, and if it helps even one other person get that โ€œaha!โ€ moment, it was worth every scribbled matrix.

Logistic Regression from A to Z.pdf (3.0 MB)

1 Like