Normalization

Normalization ensures numeric attributes are on the same scale, preventing bias in the SVM model.

SVM require the normalization of numeric input. Normalization places the values of numeric attributes on the same scale and prevents attributes with a large original scale from biasing the solution. Normalization also minimizes the likelihood of overflows and underflows.