Looking at Pearson's correlation formula and it just struck me : If we think in inner product (or dot product) terms, it looks exactly like the cosine of the angle between the two vectors representing the centered values! And the coefficient of linear regression is just the orthogonal proejction. It seems rooted in linear algebra. I'm guessing that's why the "linear" aspect is automatically built-in ?
Thank you for the videos!
Looking at Pearson's correlation formula and it just struck me : If we think in inner product (or dot product) terms, it looks exactly like the cosine of the angle between the two vectors representing the centered values!
And the coefficient of linear regression is just the orthogonal proejction. It seems rooted in linear algebra. I'm guessing that's why the "linear" aspect is automatically built-in ?
Awesome 🎉