tmadl/sklearn-interpretable-tree

Simplified tree-based classifier and regressor for interpretable machine learning (scikit-learn compatible)

sklearn-interpretable-tree's Issues

Update interpretable tree to leverage better scikit-learn's recent tree upgrades

Hello @tmadl , thank you for your work, I really appreciate it. Since I've been working a lot with explainable ML recently I found it great to discover your repositories and projects. This is why I've decided to come here to leave a suggestion.

Last (and recent) scikit-learn's version released an upgrade on their decision tree implementation, allowing for a new prunning parameter (see docs here). This is cool and I was wondering how this relates with the simplification tree strategy you've implemented. Which criteria is the best to use for prunning? Is it interesting to use both?

Depending on your views too, I would suggest then to update this tree implementation to benefit and leverage sklearn's new tree upgrade. I believe it can be not interesting too in the end. What do you think? I would love to read your comments on this.

Thank you very much.