Why represent linear-SVM and RBF-SVM searches as separate dictionaries instead of one Cartesian grid?
Classic MLLevel: IntermediateConditional parameter spaces
Why represent linear-SVM and RBF-SVM searches as separate dictionaries instead of one Cartesian grid?
Hints
- Not every parameter applies to every model family.
- A list of grids is a union, not one product.
Answer
Kernel-specific parameters such as gamma need not be explored for the linear kernel. Separate dictionaries express valid conditional combinations and avoid redundant or invalid fits.
Source and licence: scikit-learn 1.7.2 — grid search · BSD-3-Clause · answer: Quantoros editorial