EasyDeepLearn

SMOTE, Borderline-SMOTE, SVM-SMOTE, ADASYN — when do you use each?

hard

Answer

  • Vanilla SMOTE oversamples uniformly, including points deep in the class interior (redundant).
  • Borderline-SMOTE: only synthesize from minority points near the decision boundary (their kNN mixes both classes) — where the model actually needs help.
  • SVM-SMOTE: fit an SVM, synthesize near the support vectors.
  • ADASYN: synthesize more around minority points that are hard to classify (many majority neighbours), less around easy ones — adaptive.
  • In practice: try Borderline-SMOTE and ADASYN before vanilla — usually a small but consistent improvement.
Check yourself — multiple choice
  • They're all identical
  • Borderline / SVM-SMOTE focus near the decision boundary; ADASYN adapts to hard-to-classify points
  • ADASYN is only for regression
  • SVM-SMOTE doesn't use SVMs

Boundary-focused (Borderline, SVM-SMOTE) or difficulty-adaptive (ADASYN) usually beat vanilla SMOTE.

#imbalance

Practise Supervised Learning

215 interview questions in this topic.

Related questions