Diffusion Boosted Trees

TL;DR

Diffusion Boosted Trees (DBT) integrate denoising diffusion models with gradient boosting, creating a novel decision tree-based generative and predictive framework that excels in real-world regression and classification tasks, including fraud detection.

Contribution

This paper introduces DBT, a new model combining diffusion processes with boosting, offering a non-parametric approach to conditional distribution learning with practical advantages.

Findings

DBT outperforms neural diffusion models in experiments.

DBT demonstrates strong performance on real-world regression tasks.

DBT effectively applies to fraud detection with learning to defer.

Abstract

Combining the merits of both denoising diffusion probabilistic models and gradient boosting, the diffusion boosting paradigm is introduced for tackling supervised learning problems. We develop Diffusion Boosted Trees (DBT), which can be viewed as both a new denoising diffusion generative model parameterized by decision trees (one single tree for each diffusion timestep), and a new boosting algorithm that combines the weak learners into a strong learner of conditional distributions without making explicit parametric assumptions on their density forms. We demonstrate through experiments the advantages of DBT over deep neural network-based diffusion models as well as the competence of DBT on real-world regression tasks, and present a business application (fraud detection) of DBT for classification on tabular data with the ability of learning to defer.