English

Scalable Optimal Multiway-Split Decision Trees with Constraints

Machine Learning 2023-02-15 v1

Abstract

There has been a surge of interest in learning optimal decision trees using mixed-integer programs (MIP) in recent years, as heuristic-based methods do not guarantee optimality and find it challenging to incorporate constraints that are critical for many practical applications. However, existing MIP methods that build on an arc-based formulation do not scale well as the number of binary variables is in the order of O(2dN)\mathcal{O}(2^dN), where dd and NN refer to the depth of the tree and the size of the dataset. Moreover, they can only handle sample-level constraints and linear metrics. In this paper, we propose a novel path-based MIP formulation where the number of decision variables is independent of NN. We present a scalable column generation framework to solve the MIP optimally. Our framework produces a multiway-split tree which is more interpretable than the typical binary-split trees due to its shorter rules. Our method can handle nonlinear metrics such as F1 score and incorporate a broader class of constraints. We demonstrate its efficacy with extensive experiments. We present results on datasets containing up to 1,008,372 samples while existing MIP-based decision tree models do not scale well on data beyond a few thousand points. We report superior or competitive results compared to the state-of-art MIP-based methods with up to a 24X reduction in runtime.

Keywords

Cite

@article{arxiv.2302.06812,
  title  = {Scalable Optimal Multiway-Split Decision Trees with Constraints},
  author = {Shivaram Subramanian and Wei Sun},
  journal= {arXiv preprint arXiv:2302.06812},
  year   = {2023}
}
R2 v1 2026-06-28T08:39:28.928Z