Dropout layer

[lars-reimann]

Is your feature request related to a problem?

Dropout layers are a common way to regularize a neural network.

Desired solution

• Add a new class safeds.ml.nn.layers.DropoutLayer.
• Superclass: Layer
• Constructor parameters:
  • probability: float
  • Call _check_bounds to ensure that probability is between 0 and 1 (both inclusive)
• Implementation:
  • Use PyTorch's Dropout internally
  • Pass the probality along and don't work inplace.

Possible alternatives (optional)

No response

Screenshots (optional)

No response

Additional Context (optional)

No response