Maximum Margin Interval Trees

TL;DR

This paper introduces a novel nonlinear tree model for regression with interval or censored data, optimizing a margin-based objective with a dynamic programming algorithm that achieves high accuracy and speed.

Contribution

It proposes a new nonlinear tree learning method for interval regression using a margin-based objective and an efficient dynamic programming solution.

Findings

Achieves state-of-the-art prediction accuracy

Offers fast training with log-linear time complexity

Performs well on multiple benchmark datasets

Abstract

Learning a regression function using censored or interval-valued output data is an important problem in fields such as genomics and medicine. The goal is to learn a real-valued prediction function, and the training output labels indicate an interval of possible values. Whereas most existing algorithms for this task are linear models, in this paper we investigate learning nonlinear tree models. We propose to learn a tree by minimizing a margin-based discriminative objective function, and we provide a dynamic programming algorithm for computing the optimal solution in log-linear time. We show empirically that this algorithm achieves state-of-the-art speed and prediction accuracy in a benchmark of several data sets.