update.ftrlprox: Update FTRL Proximal model
Description
Continue training model on new data
Usage
"update"(object, newX, newY, num_epochs = 1, save_loss = F, ...)
Arguments
newX
new feature vectors. This needs to be the same features as used in previous training rounds for this object.
num_epochs
number of times we should traverse over the training data, defaults to 1.
save_loss
is to save the loss function during training. This will be appended to previous loss vector.
Value
ftrlprox model object
Details
As FTRL PRoximal is an online algorithm it is possible to continue training the model on new data. This can be good if for for example the size of the dataset is too large to keep in memory or new data is getting available after some time.