Is the definition of multicollinearity being features highly correlated correct? I was thinking about it and I believe it's possible to have multicollinearity without high correlation. For instance, if x1 is a result of a linear combination of x2, x3, x4, x5, x6, x7, x8, x9, x10, x11,, and each one of these features "explains" 10% of x1, they won't be highly correlated with x1, even though there is perfect multicollinearity, right?

The article is great, I don't want to bother you, but I was thinking about this particular point and I don't think I agree with it. Overall, the article is awesome, great piece of work! Congrats!!