python - Why MLARAM algorithm for multilable classification chaning the values of training labels?

Question

I am using MLARAM algorithm from skmultilearn library for doining multilabel classification. I have a set of binary labels. When I am training the MLARAM algorithm, I am providing features X_train and the binary labels Y_train for training. Once the training is completed, I am cheking the values of Y_train. Ideally, Y_train should remain binary, but I am getting a not binary categorical labels.

Why is this weired behaviour happening? Any suggestions?

Answer 1

等待大神答复