Running statsmodel for VIF( Variance Inflation factor) to check for multicollinearity is it enought

Running stats model for VIF( Variance Inflation factor) to check for multicollinearity is it enough , and can we discard variables having VIF>10 without any further confirmations

1 Like

For the purposes of detecting multi collinearity, yes, VIF is as comprehensive as it gets.

yes we can remove the value

© Copyright 2013-2020 Analytics Vidhya