Desired Output Preprocessing
In some situations, the Zero2Neuro representation for desired outputs requires one or more steps of transformation before they are in a format that can be used with the corresponding Scikit-Learn model. These transformations are implemented using a separate Pipeline that is applied just to the desired outputs.
Example
--skl_y_pipeline
TransformOutputRavel
Modules
TransformOutputRavel
This module translates desired outputs of an arbitrary shape to a single vector. Typically, this is used to translate a matrix of desired values of shape (N,1) (N examples and one value for each example) to a vector that has a shape (N,) (still N examples)
Arguments
There are no arguments for this preprocessing module.