Status:
This tutorial is much shorter and self-contained and didn't require many edits.
- Some language tweaks needed to increase readability
- "We'll close the loop on ...." (not sure I even understand what this was supposed to be).
- "Start the interactive?" - Fixed
- Sample complexity equations come out of nowhere and I don't think it's intuitive how to link the use of the interactive tool with actually collecting enough data to make that estimate that's asked about in the discussion point. (Tweaked)
- Day 2 relies heavily on the use of sample complexity (Leila's code tests on training MNIST models on increasing dataset sizes, so this part should stay to get the students used to the notion of sample complexity)
- I don't think it's clear they should be mentally implementing the binomial model just to get the basic point across though. In the end, I've left this as it was but am not so happy with its inclusion