Searcharxiv⌕ Search

arXiv subjects

Sihan Wu

Publications and source records attributed to Sihan Wu.

5 recordsLinked to original sources

Learning the Average Dose-Response Function with a Right-Censored Treatment

This paper considers estimation and inference for the average dose-response function under right-censored treatment duration and missing outcomes for censored units. Using the duration lower bounds provided by censored records, we construct a novel multiply robust estimator through augmented local estimating equations. Under suitable regularity conditions, the estimator remains consistent when any two of the outcome regression, treatment density, and censoring survival function are consistently estimated. We establish its convergence rate and asymptotic distribution within a cross-fitting framework under appropriate conditions on nuisance estimation. To address local smoothing bias that can distort confidence interval coverage, we further develop a debiased estimator that supports asymptotically valid pointwise inference without undersmoothing. Under additional uniform conditions, we construct simultaneous confidence bands using a Gaussian multiplier bootstrap. Simulation studies evaluate estimation accuracy, the validity of pointwise and uniform inference, and robustness to nuisance model misspecification. A semi-synthetic study calibrated to NHANES illustrates the proposed methods in a setting with a known dose-response function.

math.ST↗

Debiased inference for proximal dose-response function

We study nonparametric inference for a continuous-treatment causal dose-response curve and its first derivative under unmeasured confounding, using treatment- and outcome-inducing proxies. We establish a proximal doubly robust pseudo-outcome whose conditional mean given treatment equals the dose-response curve whenever either bridge function is correctly specified. Their (smoothed) efficient influence function representations lead to cross-fitted local-polynomial estimators with explicit smoothing-bias correction. We establish pointwise and finite-dimensional asymptotic normality and oracle uniform Gaussian approximations over compact interior treatment intervals, allowing mean-squared-error-optimal bandwidths without undersmoothing. The derivative inference reuses the estimated bridges and the same form of bandwidth-dependent nuisance-rate conditions, evaluated at the bandwidth rate appropriate for derivative estimation, without requiring derivative-specific bridge estimation. Employing cross-fitting accommodates flexible bridge estimators under product convergence rate conditions without entropy restrictions. We further develop tests against fixed-dimensional parametric dose-response null hypotheses, including constancy, linearity, and quadratic specifications, and derive their limiting distributions under the null and local alternatives, together with practical bootstrap procedures. We also develop practical bandwidth selectors, pointwise confidence intervals, and simultaneous confidence bands for each causal estimand. Extensive simulations and a data analysis, presented in the appendix, demonstrate the finite-sample performance of the proposed methods under latent confounding and multiple proxies.

stat.ME↗

Proximal Mediation Analysis with Hidden Recanting Witnesses

Mediation analysis is essential for decomposing the causal effect of a treatment into direct and indirect pathways. However, many practical settings rely on the stringent assumption that recanting witnesses, defined as treatment-induced mediator-outcome confounders, are either absent or fully known a priori. Such a requirement is often untenable, especially when these variables remain unobservable due to measurement difficulties or privacy constraints. In this paper, we leverage proximal causal inference to develop three novel identification strategies to address the challenge of identifying path-specific effects in the presence of unknown recanting witnesses. Building on this, we develop a semiparametric inference framework that derives the efficient influence function and proposes a proximal multiply robust estimator, which remains consistent if at least one set of nuisance models is correctly specified. When all nuisance models are correctly specified and converge at appropriate rates, the estimator is asymptotically normal and achieves the semiparametric efficiency bound. We provide a minimax optimization-based debiased machine learning procedure for point estimation and constructing valid confidence intervals. The performance of the proposed methods is demonstrated by simulation studies and a real data application.

stat.ME↗

Proximal Path-Specific Inference

Causal mediation analysis has been extended to estimate path-specific effects with multiple intermediate variables, isolating treatment effects through a mediator of interest while excluding pathways through its ancestors. Such analyses address bias from recanting witnesses, i.e., treatment-induced mediator-outcome confounders. However, existing methods typically rely on stringent assumptions precluding general unmeasured confounding, which are often violated in practice. In this paper, we relax these restrictions by leveraging observed covariates as proxy variables to accommodate unmeasured confounding among the treatment, recanting witness, mediator, and outcome. Using proximal confounding bridge functions, we develop four nonparametric identification strategies for the path-specific effect. We further derive the efficient influence function and propose a quadruply robust, locally efficient estimator. To handle high-dimensional nuisance parameters, we propose a proximal debiased machine learning approach. We theoretically guarantee that our estimator achieves $\sqrt{n}$-consistency and asymptotic normality even when machine learning estimators for nuisance functions converge at slower rates. Our approaches are validated via semiparametric and nonparametric simulations and an application to the CDC WONDER Natality study, estimating the path-specific effect of prenatal care on preterm birth through preeclampsia, independent of maternal smoking during pregnancy.

stat.ME↗

Do Transformers Have the Ability for Periodicity Generalization?

Large language models (LLMs) based on the Transformer have demonstrated strong performance across diverse tasks. However, current models still exhibit substantial limitations in out-of-distribution (OOD) generalization compared with humans. We investigate this gap through periodicity, one of the basic OOD scenarios. Periodicity captures invariance amid variation. Periodicity generalization represents a model's ability to extract periodic patterns from training data and generalize to OOD scenarios. We introduce a unified interpretation of periodicity from the perspective of abstract algebra and reasoning, including both single and composite periodicity, to explain why Transformers struggle to generalize periodicity. Then we construct Coper about composite periodicity, a controllable generative benchmark with two OOD settings, Hollow and Extrapolation. Experiments reveal that periodicity generalization in Transformers is limited, where models can memorize periodic data during training, but cannot generalize to unseen composite periodicity. We release the source code to support future research.

cs.LG↗