Explain 'Soft Margin' in SVM.

Explain 'Soft Margin' in SVM.

Approach: Describe allowing some misclassifications to improve overall generalization.

An extension of SVM that allows for some misclassification in exchange for a wider margin. The 'C' hyperparameter controls this trade-off: small C allows more misclassifications (wider margin), large C penalizes them heavily (narrower margin).

Follow-up: What happens if C is infinity?

Key concepts: slack variables, c parameter, overfitting.