It is available to design the loss function by myself? If so, how can i change the loss function? about deepctr HOT 2 CLOSED

Hi, @Jimmy880
You can use the class_weight arg in fit method.
for example,if we want to set the pos_weight=1.5 and neg_weight=1,then

model.fit(x,y,class_weight={0:1.0,1:1.5})

from deepctr.

@shenweichen
Thanks for your reply. It really helps me a lot in training imbalance dataset.

I wonder whether the class_weight parameter should be passed in the dict format. I used xgboost and sklearn before and in their lib the weight of binary class can be passed in a ratio as negative sample number/ positive sample number. Could it also work in here?

from deepctr.