What is VIF and when do you worry about multicollinearity?
mediumAnswer
- Variance Inflation Factor of feature , where is the of regressing on the other features.
- VIF = 1 means no collinearity; VIF > 5 or 10 is a common warning threshold.
- High VIF inflates coefficient standard errors, makes individual coefficients unstable and hard to interpret, but doesn't harm overall prediction.
- Fixes: drop redundant features, combine them (PCA), or use ridge regression which is stable under collinearity.
Check yourself — multiple choice
- Low VIF (<0.5) signals multicollinearity
- High VIF (>5–10) signals multicollinearity and unstable coefficients
- VIF measures overfitting
- VIF is only used for classification
VIF > 5–10 flags multicollinearity — coefficients become unstable but predictions remain OK.
#linear-regression#assumptions#features
Practise Supervised Learning
215 interview questions in this topic.
Related questions
- What are the classical assumptions of linear regression?
- How do you handle strong multicollinearity in a linear model?
- How do you detect and fix heteroscedasticity?
- What's the main risk of polynomial regression?
- When should you add interaction features to a linear model?
- Which models need feature scaling and which don't?