EasyDeepLearn

Why is a single decision tree considered a 'high-variance' model?

medium

Answer

  • A small change in the training data can produce a completely different tree: the first split cascades, so if a different feature is chosen at the top, everything below changes.
  • This makes individual trees unstable — predictions vary a lot with resampling.
  • Bagging (Random Forest) and boosting exploit this: averaging many independent (or corrected-residual) trees reduces variance dramatically without hurting bias.
Check yourself — multiple choice
  • Trees have low variance — that's why we use them
  • Small data changes flip the top splits and cascade through the whole tree
  • Trees always have zero training error
  • Trees are stable but slow

Top-split cascades ⇒ high variance ⇒ perfect candidate for bagging / boosting.

#decision-trees#trees#theory

Practise Supervised Learning

215 interview questions in this topic.

Related questions