Robust Watermarking on Gradient Boosting Decision Trees

2025-11-12

AI Summary

Gradient Boosting Decision Trees (GBDTs) lack robust watermarking techniques for intellectual property protection. Method: This paper proposes the first in-situ robust watermarking framework specifically designed for GBDTs. It embeds watermarks by leveraging tree-structure-aware node-split path encoding and imperceptible weight perturbations—without altering the model architecture—and introduces four embedding strategies to balance fidelity and robustness. Contribution/Results: The method maintains high model accuracy (accuracy degradation <1%) while significantly enhancing resilience against post-deployment fine-tuning and other adversarial attacks. Extensive experiments on multiple benchmark datasets demonstrate high watermark extraction rates (>98%) and strong robustness under diverse perturbations, effectively addressing a critical gap in GBDT copyright protection. This work establishes a novel paradigm for securing traditional machine learning models, advancing practical IP protection for non-neural ML systems.

Technology Category

• Machine Learning: Adversarial Learning & Robustness
• Computer Vision: Adversarial Attacks & Robustness
• Natural Language Processing: Safety and Robustness

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Gradient Boosting Decision Trees (GBDTs) are widely used in industry and academia for their high accuracy and efficiency, particularly on structured data. However, watermarking GBDT models remains underexplored compared to neural networks. In this work, we present the first robust watermarking framework tailored to GBDT models, utilizing in-place fine-tuning to embed imperceptible and resilient watermarks. We propose four embedding strategies, each designed to minimize impact on model accuracy while ensuring watermark robustness. Through experiments across diverse datasets, we demonstrate that our methods achieve high watermark embedding rates, low accuracy degradation, and strong resistance to post-deployment fine-tuning.

Problem

• Develop robust watermarking framework for Gradient Boosting Decision Trees
• Embed imperceptible watermarks while maintaining model accuracy
• Ensure watermark resilience against post-deployment fine-tuning attacks

Innovation

• Robust watermarking framework for GBDT models
• In-place fine-tuning embeds resilient watermarks
• Four embedding strategies maintain model accuracy