Verifying Robustness of Gradient Boosted Models

TL;DR

This paper introduces VeriGB, a novel verification tool that encodes gradient boosted models as SMT formulas to rigorously assess their robustness against input perturbations, addressing a key gap in model reliability evaluation.

Contribution

The paper presents VeriGB, the first method to formally verify the robustness of gradient boosted models using SMT encoding, enabling analysis of large models and comparison of configurations.

Findings

VeriGB can verify large gradient boosted models effectively.

Some model configurations are inherently more robust.

VeriGB outperforms existing approaches in robustness verification.

Abstract

Gradient boosted models are a fundamental machine learning technique. Robustness to small perturbations of the input is an important quality measure for machine learning models, but the literature lacks a method to prove the robustness of gradient boosted models. This work introduces VeriGB, a tool for quantifying the robustness of gradient boosted models. VeriGB encodes the model and the robustness property as an SMT formula, which enables state of the art verification tools to prove the model's robustness. We extensively evaluate VeriGB on publicly available datasets and demonstrate a capability for verifying large models. Finally, we show that some model configurations tend to be inherently more robust than others.