Kernel methods for causal functions: dose, heterogeneous and incremental response curves

• 成果类型: Article
• 署名作者: Singh, R.; Xu, L.; Gretton, A.
• 署名单位: Massachusetts Institute of Technology (MIT); University of London; University College London
• 刊物名称: BIOMETRIKA
• ISSN/ISSBN: 0006-3444
• DOI: 10.1093/biomet/asad042
• 发表日期: 2024
• 页码: 497516
• 关键词: efficient semiparametric estimation; inference; variance

摘要

We propose estimators based on kernel ridge regression for nonparametric causal functions such as dose, heterogeneous and incremental response curves. The treatment and covariates may be discrete or continuous in general spaces. Because of a decomposition property specific to the reproducing kernel Hilbert space, our estimators have simple closed-form solutions. We prove uniform consistency with finite sample rates via an original analysis of generalized kernel ridge regression. We extend our main results to counterfactual distributions and to causal functions identified by front and back door criteria. We achieve state-of-the-art performance in nonlinear simulations with many covariates, and conduct a policy evaluation of the US Job Corps training programme for disadvantaged youths.