EasyDeepLearn

What's the main risk of polynomial regression?

easy

Answer

  • High-degree polynomial features overfit dramatically: they can hit training points exactly while oscillating wildly between them (Runge's phenomenon).
  • Predictions extrapolate very poorly outside the training range.
  • Alternatives: use splines (piecewise low-degree polynomials with continuity), local polynomials (LOESS), regularized polynomials, or non-parametric models like trees.
  • If you must use polynomials, standardize inputs and combine with L2 regularization.
Check yourself — multiple choice
  • It's always worse than linear regression
  • High-degree polynomials overfit and extrapolate badly (Runge's phenomenon)
  • Polynomials can't be regularized
  • It works only for one feature

High-degree polynomials overfit and oscillate; splines or regularized fits are usually safer.

#linear-regression#features#regularization

Practise Supervised Learning

215 interview questions in this topic.

Related questions