Summer of Math Exposition

Hi all, I've written an interactive blog post with a basic introduction to prediction in Machine Learning. This is a topic that I covered in depth somewhat-recently in my masters degree and I feel like the textbooks didn't work very hard to provide the "click" moment. After the click came, all of the maths made much more sense. I think that videos and interactive figures are a great way of bringing across this intuition. I'm really trying to reinforce the flexibility. You don't have to use one of the standard models, like neural networks, gradient boosting, linear models etc. It is possible to make up your own model that fits your problem and then figure out how to train/optimise it. The models that we have exist for a reason. They're good all-rounders. We know how to optimise them. But often, especially when you don't have much data, making something bespoke is the way to go. I've also tried to emulate Grant's approach to teaching, where the learner should feel like they've discovered something themselves along the way. I want to introduce the problem of prediction as something you *could* do by hand, but that it's better and easier to pass it over to an algorithm. I've also tried to keep the formal and informal parts separate. There's more background in the footnotes for those that would like to have a more structured description of the ideas. I'm not a teacher, so one of the hardest parts was imagining my target audience and what they might struggle with. I believe that this might be useful for advanced high school students and early undergrads. Any feedback on anything that is too easy or too hard given this audience would be much appreciated! Thanks to you all for your part in making maths a little more beautiful and intuitive! Cheers, Jason