TY - GEN
T1 - SCI
T2 - 30th ACM International Conference on Information and Knowledge Management, CIKM 2021
AU - Yao, Liuyi
AU - Li, Yaliang
AU - Li, Sheng
AU - Huai, Mengdi
AU - Gao, Jing
AU - Zhang, Aidong
N1 - Publisher Copyright:
© 2021 ACM.
PY - 2021/10/30
Y1 - 2021/10/30
N2 - Inferring causal effect from observational data has attracted much attention from various domains. Under the potential outcome framework, the estimation of counterfactuals is crucial for the investigation of causal effect at the individual level. Existing representation learning approaches focus on learning one balanced feature space, which ignores certain information predictive to the outcomes. To fully utilize the predictive information, we propose a Subspace learning based Counterfactual Inference (SCI) method to estimate causal effect at the individual level. Different from existing work, SCI learns both a common subspace, which preserves the information across all the treatment groups, and treatment-specific subspaces, which retain the information associated with each specific treatment. Learning from two kinds of subspaces helps SCI obtain better causal effect estimations than state-of-the-art methods, demonstrated by a series of experiments on synthetic and real-world datasets.
AB - Inferring causal effect from observational data has attracted much attention from various domains. Under the potential outcome framework, the estimation of counterfactuals is crucial for the investigation of causal effect at the individual level. Existing representation learning approaches focus on learning one balanced feature space, which ignores certain information predictive to the outcomes. To fully utilize the predictive information, we propose a Subspace learning based Counterfactual Inference (SCI) method to estimate causal effect at the individual level. Different from existing work, SCI learns both a common subspace, which preserves the information across all the treatment groups, and treatment-specific subspaces, which retain the information associated with each specific treatment. Learning from two kinds of subspaces helps SCI obtain better causal effect estimations than state-of-the-art methods, demonstrated by a series of experiments on synthetic and real-world datasets.
KW - causal inference
KW - treatment effect estimation
UR - https://www.scopus.com/pages/publications/85119170987
U2 - 10.1145/3459637.3482175
DO - 10.1145/3459637.3482175
M3 - Conference contribution
AN - SCOPUS:85119170987
T3 - International Conference on Information and Knowledge Management, Proceedings
SP - 3583
EP - 3587
BT - CIKM 2021 - Proceedings of the 30th ACM International Conference on Information and Knowledge Management
PB - Association for Computing Machinery
Y2 - 1 November 2021 through 5 November 2021
ER -