You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In the old implementation of autopytorch, there is fast ICA as a feature preprocessor. However, due to a bug in scikit learn's implementation causes numerical instability and there crashes.
Also, fast ica is not a typical dimensionality reduction.
Therefore, we should analyse if it makes sense to add fast ICA component to the new version
The text was updated successfully, but these errors were encountered:
In the old implementation of autopytorch, there is fast ICA as a feature preprocessor. However, due to a bug in scikit learn's implementation causes numerical instability and there crashes.
Also, fast ica is not a typical dimensionality reduction.
Therefore, we should analyse if it makes sense to add fast ICA component to the new version
The text was updated successfully, but these errors were encountered: