Causal Effect Estimation: Recent Advances, Challenges, and Opportunities

• DOI: 10.48550/arxiv.2302.00848
• Publication Date: 2023-01-01
• Authors: Zhixuan Chu; Jianmin Huang; Ruopeng Li; Wei Chu; Sheng Li
• Keywords: Methodology (stat.ME); FOS: Computer and information sciences; Computer Science - Machine Learning; Statistics - Machine Learning; Machine Learning (stat.ML); 0101 mathematics; 01 natural sciences; Statistics - Methodology; Machine Learning (cs.LG)

Causal inference has numerous real-world applications in many domains, such as health care, marketing, political science, and online advertising. Treatment effect estimation, a fundamental problem causal inference, been extensively studied statistics for decades. However, traditional treatment estimation methods may not well handle large-scale high-dimensional heterogeneous data. In recent years, an emerging research direction attracted increasing attention the broad artificial intelligence field, which combines advantages of approaches (e.g., propensity score, matching, reweighing) advanced machine learning representation learning, adversarial graph neural networks). Although have shown extraordinary performance it also comes with lot new topics questions. view latest efforts we provide comprehensive discussion challenges opportunities three core components task, i.e., treatment, covariates, outcome. addition, showcase promising directions this topic from multiple perspectives.