Thanks for your comment and suggestion. I understand that math makes it far less approachable, but I did it like that based on my experience. I understand the concept easy but I was missing the explanation of backprop and all those formulas. However, I would reconsider putting more about the intuition of NN (for broader audience) and maybe making the math part more like "optional" or "bonus" part of the post.
I mean, to each their own. It’s probably better we (as a community) present it both ways. I just typically find people interested in neural networks (data scientists or the like) use the mathematical representations, but engineers typically just have their eyes glaze over. I suppose it depends on the target audience
