Set kernel of `SupportVectorMachine`

[lars-reimann]

Is your feature request related to a problem?

It's not possible to set the kernel of a SupportVectorMachine.

Desired solution

• Add a new parameter kernel: SupportVectorMachineKernel to the initializer of safeds.ml.classification.SupportVectorMachine and safeds.ml.regression.SupportVectorMachine
• Pass it along as the kernel of the wrapped scikit-learn model in the fit method

The implementation of SupportVectorMachineKernel should be similar to the ImputerStrategy:

• SupportVectorMachineKernel should be an abstract base class
• Kernel should be nested into SupportVectorMachine
• Subclasses should be nested into Kernel
• Subclasses should be:
  • Linear
  • Polynomial (with a degree: int parameter)
    • Raise if degree < 1
  • Sigmoid
  • RadialBasisFunction

Also add a getter as described in #260.

Possible alternatives (optional)

No response

Screenshots (optional)

No response

Additional Context (optional)

We need to add the support vector machine first (#154).