Enable training with weighted/scaled data points. The weight vector s is an optional parameter that holds the weight of each data point in X. s is either None or an np.array with shape (n_samples,). If s is used then the loss function is either loss_weighted_logistic or loss_weighted_mse.
In order to use the weights then use_weights=True has to be set when initializing the TFFMClassifier or TFFMRegressor.
Enable training with weighted/scaled data points. The weight vector
sis an optional parameter that holds the weight of each data point inX.sis eitherNoneor annp.arraywith shape(n_samples,). Ifsis used then the loss function is eitherloss_weighted_logisticorloss_weighted_mse.In order to use the weights then
use_weights=Truehas to be set when initializing theTFFMClassifierorTFFMRegressor.@geffy