to use linear regression we should have the predictors/features be independent from one another. However, in polynomial regression, x^2 is perfectly correlated to x ( knowing x value gives us the exact value of x^2 and vice versa ). So isn't that a problem for regression ? because ridge regression tries to solve this precise issue.
Asked
Active
Viewed 34 times
0
-
1"Correlated with" is not the same thing as "is a function of" or "can be predicted by." That's worth bearing in mind when you review the duplicate threads. – whuber Dec 20 '21 at 01:57