One-Class SVM

Margin

Distance between the decision hyperplane and the closest data points (support vectors), maximized to improve the robustness of the SVM model.

← Indietro