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
For mini-batch gradient update, we need to shuffle training data, so that actual training inputs for each iteration are different and random. This should benefit model performance given same iterations of training.
Applicable models include NN, LR and WDL.
The text was updated successfully, but these errors were encountered:
For mini-batch gradient update, we need to shuffle training data, so that actual training inputs for each iteration are different and random. This should benefit model performance given same iterations of training.
Applicable models include NN, LR and WDL.
The text was updated successfully, but these errors were encountered: