midSVM

What is gamma in SVM?

Updated May 17, 2026

Short answer

Gamma defines how far the influence of a single training example reaches in RBF kernel.

Deep explanation

A small gamma means far influence and smoother decision boundary, while large gamma leads to tighter, more complex boundaries that may overfit.

Real-world example

Used in fraud detection systems for tuning model sensitivity.

Common mistakes

  • Setting gamma too high and causing overfitting.

Follow-up questions

  • How does gamma affect bias-variance?
  • What happens if gamma is too small?

More SVM interview questions

View all →