This Julia package implements our unified framework of quantification and unfolding algorithms.
A detailed documentation is available online.
using QUnfold, ScikitLearn
@sk_import ensemble: RandomForestClassifier
# X_trn, y_trn = my_training_data(...)
acc = ACC( # a scikit-learn classifier is needed
RandomForestClassifier(oob_score=true) # we recommend bagging with oob_score=true
)
trained_acc = fit(acc, X_trn, y_trn) # fit returns a trained COPY
# X_tst = my_testing_data(...)
p_est = predict(trained_acc, X_tst) # return a prevalence vector