bsnsing: A decision tree induction method based on recursive optimal boolean rule composition

TL;DR

This paper introduces bsnsing, a novel decision tree induction method using a new MIP formulation that directly maximizes Gini reduction, offering faster training and better discrimination on benchmark datasets.

Contribution

The paper presents a new MIP-based decision tree method that optimizes split rules directly for Gini reduction and retains recursive partitioning flexibility, improving speed and accuracy.

Findings

Outperforms other decision tree algorithms in discrimination ability.

Achieves faster training times compared to existing optimal decision tree methods.

Maintains high prediction accuracy across diverse datasets.

Abstract

This paper proposes a new mixed-integer programming (MIP) formulation to optimize split rule selection in the decision tree induction process, and develops an efficient search algorithm that is able to solve practical instances of the MIP model faster than commercial solvers. The formulation is novel for it directly maximizes the Gini reduction, an effective split selection criterion which has never been modeled in a mathematical program for its nonconvexity. The proposed approach differs from other optimal classification tree models in that it does not attempt to optimize the whole tree, therefore the flexibility of the recursive partitioning scheme is retained and the optimization model is more amenable. The approach is implemented in an open-source R package named bsnsing. Benchmarking experiments on 75 open data sets suggest that bsnsing trees are the most capable of discriminating new cases compared to trees trained by other decision tree codes including the rpart, C50, party and tree packages in R. Compared to other optimal decision tree packages, including DL8.5, OSDT, GOSDT and indirectly more, bsnsing stands out in its training speed, ease of use and broader applicability without losing in prediction accuracy.