Learning covariate importance for matching in policy-relevant observational research

Hongzhe Zhang, Jiasheng Shi, Jing Huang

Matching methods are widely used to reduce confounding effects in observational studies, but conventional approaches often treat all covariates as equally important, which can resu…