End-to-end Learning of Deterministic Decision Trees

Thomas Hehn, Fred A. Hamprecht

Conventional decision trees have a number of favorable properties, including interpretability, a small computational footprint and the ability to learn from little training data. H…