Having trained N trees, you can then treat each of those N predictions as a feature itself, and run a logistic regression over that (thus "stacking" a logistic regression model on top of the decision tree model). It would be interesting to experiment with this.
Having trained N trees, you can then treat each of those N predictions as a feature itself, and run a logistic regression over that (thus "stacking" a logistic regression model on top of the decision tree model). It would be interesting to experiment with this.