How does the C parameter in an SVM affect the fit?
mediumAnswer
- C weights the misclassification penalty in the soft-margin loss (opposite of a typical regularization strength — larger C means less regularization).
- Large C: try hard to classify every training point correctly → narrow margin, complex boundary, high variance.
- Small C: tolerate more margin violations → wider margin, simpler boundary, high bias.
- Tune C on a log-spaced grid with CV.
- With an RBF kernel, C interacts with gamma — tune them jointly.
Check yourself — multiple choice
- Large C = more regularization
- Large C = less regularization / narrow margin; small C = wider margin, more bias
- C only affects training speed
- C is fixed at 1 in scikit-learn
C is inversely a regularization strength — large C ⇒ hard-margin-like ⇒ overfits.
#svm#hyperparameter-tuning#regularization
Practise Supervised Learning
215 interview questions in this topic.