ImplementBalancedRandomForestClassifier

In Python, I am used to using the imbalanced-learn library for imbalanced datasets. That library simply builds upon scikit-learn. I am particularly interested in the BalancedRandomForestClassifier.

This is a random forest that only resamples the majority class, so that all decision trees of the forest are trained on balanced bootstrapped samples of the original dataset.

Is there any plan to implement this algorithm in smile? otherwise could it be added to the roadmap? I would potentially be interested in contributing if you think that would be relevant to add this algorithm to smile.

More generally, the BalancedBaggingClassifier could also be implemented.

References:
https://imbalanced-learn.org/stable/generated/imblearn.ensemble.BalancedBaggingClassifier.html
https://imbalanced-learn.org/stable/generated/imblearn.ensemble.BalancedRandomForestClassifier.html
https://statistics.berkeley.edu/sites/default/files/tech-reports/666.pdf

该提问来源于开源项目：haifengl/smile