Abstract:Causal effect estimation under networked interference is an important but challenging problem. Available parametric methods are limited in their model space, while previous semiparametric methods, e.g., leveraging neural networks to fit only one single nuisance function, may still encounter misspecification problems under networked interference without appropriate assumptions on the data generation process. To mitigate bias stemming from misspecification, we propose a novel doubly robust causal effect estimator under networked interference, by adapting the targeted learning technique to the training of neural networks. Specifically, we generalize the targeted learning technique into the networked interference setting and establish the condition under which an estimator achieves double robustness. Based on the condition, we devise an end-to-end causal effect estimator by transforming the identified theoretical condition into a targeted loss. Moreover, we provide a theoretical analysis of our designed estimator, revealing a faster convergence rate compared to a single nuisance model. Extensive experimental results on two real-world networks with semisynthetic data demonstrate the effectiveness of our proposed estimators.

What problem does this paper attempt to address?

The paper primarily addresses the issue of causal effect estimation under network interference and proposes a novel doubly robust causal effect estimation method. Specifically, the paper aims to solve the following key problems: 1. **Challenges of Causal Effect Estimation under Network Interference**: In a network environment, interactions between individuals can lead to the failure of the traditional Stable Unit Treatment Value Assumption (SUTVA), thereby complicating causal inference. Existing parametric methods are limited by model space, while semi-parametric methods (such as using neural networks to fit a single interference function) may encounter model misspecification issues without appropriate data generation process assumptions. 2. **Reducing Bias Caused by Model Misspecification**: To mitigate the estimation bias caused by model misspecification, the paper proposes a new doubly robust causal effect estimator. This estimator trains neural networks by adapting target learning techniques and achieves doubly robustness under specific conditions. 3. **Designing Target Estimators with Lower Bias and Doubly Robustness**: The authors propose an end-to-end causal effect estimator named TNet, which ensures that the estimator maintains doubly robust properties in the presence of network interference by translating theoretical conditions into target loss functions. 4. **Convergence Rate Analysis**: The paper also provides a theoretical analysis of the designed estimator, revealing its faster convergence rate compared to single interference models. 5. **Experimental Validation**: Extensive experiments on semi-synthetic data from two real-world networks validate the effectiveness of the proposed estimator and the correctness of the theoretical analysis. In summary, the goal of this paper is to develop a novel doubly robust causal effect estimation method in the context of network interference to overcome the limitations and challenges of existing methods. By combining target learning techniques and neural networks, the proposed method aims to improve estimation accuracy and reduce bias caused by model misspecification.

Doubly Robust Causal Effect Estimation under Networked Interference via Targeted Learning

Nonparametric Doubly Robust Estimation of Causal Effect on Networks in Observational Studies

Causal effect estimation under network interference with mean-field methods

Online Experimental Design With Estimation-Regret Trade-off Under Network Interference

Graph Machine Learning based Doubly Robust Estimator for Network Causal Effects

Causal Inference under Networked Interference and Intervention Policy Enhancement

Integrating Active Learning in Causal Inference with Interference: A Novel Approach in Online Experiments

Toward Controlling Hidden Confounding for Causal Effect Estimation with Networked Observational Data.

Learning Individual Causal Effects from Networked Observational Data

Deconfounding with Networked Observational Data in a Dynamic Environment

Learning Individual Treatment Effects from Networked Observational Data.

Estimation of individual causal effects in network setup for multiple treatments

A systematic investigation of classical causal inference strategies under mis-specification due to network interference

Causal Inference Under Network Interference: A Framework for Experiments on Social Networks

Learning Individual Treatment Effects under Heterogeneous Interference in Networks.

Estimation of Causal Effects Under K-Nearest Neighbors Interference

Modeling Interference for Individual Treatment Effect Estimation from Networked Observational Data

Estimating Peer Direct and Indirect Effects in Observational Network Data

Causal Inference in Social Platforms Under Approximate Interference Networks

Treatment Effect Estimation Amidst Dynamic Network Interference in Online Gaming Experiments

Just Ramp-up: Unleash the Potential of Regression-based Estimator for A/B Tests under Network Interference