What is the role of normalization in kernel SVM?

Normalization ensures all features contribute equally to kernel similarity.

Kernel functions like RBF rely on distance computations. Without normalization, features with large scales dominate similarity measures, distorting decision boundaries and reducing performance.