Why pruning is not currently supported in scikit-learn? How can we tune the decision trees to make a workaround?
This is something which is planned to be done. Setting the minimum number of samples required at a leaf node or a split as well as setting the maximum depth of the tree are how you want to work around this.
Have a look at the 0.22 dev version of sklearn. Looks like tree pruning will be implemented in the next version.
https://scikit-learn.org/dev/whats_new.html section sklearn.tree