A Comparative Study of Model Adaptation Strategies for Multi-Treatment Uplift Modeling
Ruyue Zhang, Xiaopeng Ke, Ming Liu, Fangzhou Shi, Chang Men, Zhengdan Zhu
TL;DR
This work addresses multi-treatment uplift modeling by classifying existing adaptations into Feature Adaptation and Structure Adaptation, and introducing Orthogonal Function Adaptation (OFA) as a plug-and-play module grounded in the Weierstrass approximation. The method uses an orthogonal polynomial basis (Legendre) to model treatment effects with coefficients conditioned on hidden features, and a loss that combines BCE with distribution-discrepancy terms to correct for selection bias. Through extensive experiments on synthetic and real-world data, OFA consistently outperforms vanilla adaptation baselines and demonstrates strong robustness to noise, bias, and complex treatment-response functions, while remaining compatible with multiple underlying models. The results suggest OFA offers a practical, generalizable boost for multi-treatment causal effect estimation in domains such as marketing and healthcare.
Abstract
Uplift modeling has emerged as a crucial technique for individualized treatment effect estimation, particularly in fields such as marketing and healthcare. Modeling uplift effects in multi-treatment scenarios plays a key role in real-world applications. Current techniques for modeling multi-treatment uplift are typically adapted from binary-treatment works. In this paper, we investigate and categorize all current model adaptations into two types: Structure Adaptation and Feature Adaptation. Through our empirical experiments, we find that these two adaptation types cannot maintain effectiveness under various data characteristics (noisy data, mixed with observational data, etc.). To enhance estimation ability and robustness, we propose Orthogonal Function Adaptation (OFA) based on the function approximation theorem. We conduct comprehensive experiments with multiple data characteristics to study the effectiveness and robustness of all model adaptation techniques. Our experimental results demonstrate that our proposed OFA can significantly improve uplift model performance compared to other vanilla adaptation methods and exhibits the highest robustness.