This is a package that takes in X and y as training data and predicts two outputs as y1 and y2 for any given test data. This package removes the limitation of just have one target feature in machine learning.
Contributions are always welcome!
The package needs to be extend to predict non-categorical variables.
Please adhere to this project's code of conduct
.
To install this package run
pip install multilabeler
If you have any feedback, please reach out to me at owodunniabraham@gmail.com
from multilabeler import BilableClassifier
model = BilableClassifier(xgboost())