Outlier-Resistant Heterogeneous Treatment Effect Estimation in HDLSS Settings via GAT--CVAE Framework

2025/09/13 by Byeonghee Lee, Lee, Byeonghee, Joonsung Kang +1
Engineering · #Autoencoder #Cluster analysis #Estimation #FOS: Computer and information sciences #Fault Detection and Control Systems #Graph #Methodology (stat.ME) #Outlier #Representation (politics) #Sample (material) #Sample size determination

paper · pdf · doi:10.48550/arxiv.2509.10787

published in arXiv (Cornell University) (Cornell University)

openalex publication_date 2025/09/13 · openalex created_date 2025/10/12 · openalex updated_date 2026/08/05

Abstract

We introduce a robust framework for heterogeneous treatment effect (HTE) estimation tailored to high-dimensional low sample size (HDLSS) settings. By combining Graph Attention Networks (GAT) to capture structural dependencies among confounders with a Conditional Variational Autoencoder (CVAE) for latent representation learning, our method expands the sample space and performs clustering that integrates even outlier sets into coherent subgroups. Clusterwise causal effects are then estimated using a doubly robust outlier-resistant estimator, yielding stable and generalizable results. Simulations and real-world applications confirm superior performance compared with existing HTE methods, highlighting the framework's potential for precision medicine and policy evaluation.

Citations

• Adapting Neural Networks for the Estimation of Treatment Effects
• Graph Attention Networks
• Estimation and Inference of Heterogeneous Treatment Effects using Random Forests
• Estimation and Inference of Heterogeneous Treatment Effects using Random Forests
• Survey of Clustering Algorithms
• A Theorem on the Manipulability of Redundant Serial Kinematic Chains
• The central role of the propensity score in observational studies for causal effects

Related

• Improving Data-driven Heterogeneous Treatment Effect Estimation Under Structure Uncertainty
• Heterogeneous Treatment Effect Estimation through Deep Learning
• Estimating and discovering heterogeneous treatment effects using machine learning in epidemiological studies: a practical guide
• Causal Rule Forest: Toward Interpretable and Precise Treatment Effect Estimation
• Stable Heterogeneous Treatment Effect Estimation across Out-of-Distribution Populations