Algorithm-aware Statistical Inference in High-dimensional Estimating Equations
Abstract
Iterative algorithms are integral to many statistical procedures, whereas sampling uncertainty along their iteration is generally not quantified. We develop simultaneous inference for discrete trajectories generated from high-dimensional estimating equations, with iteration-specific population estimands that preserve algorithmic regularization. Our proposed fixed-Jacobian bootstrap propagates perturbations along the observed trajectory without repeatedly running the original nonlinear algorithm. We establish uniform first-order expansions and bootstrap validity over possibly infinite sets of contrasts and iterations under stability, entropy and moment conditions. Simulations in sparse linear regression, paired Fisher discrimination, and temporal-difference learning demonstrate accurate simultaneous coverage. A micro-randomized trial application illustrates inference for a treatment trajectory with participant-level clustering.
Related
- High-Dimensional Statistical Inference for Sparse Support Vector Machines
- Statistical inference for rough Burgers-type stochastic partial differential equations
- Statistical Methods for Estimating Probability of Detection in Structural Health Monitoring
- Semiparametrically efficient semi-supervised statistical inference
- Statistical inference and applications of Kumaraswamy Odd Perks power Lomax distribution
BibTeX
@misc{li2026algorithmaware,
title = {{Algorithm-aware Statistical Inference in High-dimensional Estimating Equations}},
author = {Tongyu Li and Alexander Giessing},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.12287},
eprint = {2611.12287},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.12287},
}
RIS
TY - UNPB
TI - Algorithm-aware Statistical Inference in High-dimensional Estimating Equations
AU - Tongyu Li
AU - Alexander Giessing
PY - 2026
DA - 2026/11/09
PB - arXiv
DO - 10.48550/arxiv.2611.12287
UR - https://arxiv.org/abs/2611.12287
AB - Iterative algorithms are integral to many statistical procedures, whereas sampling uncertainty along their iteration is generally not quantified. We develop simultaneous inference for discrete trajectories generated from high-dimensional estimating equations, with iteration-specific population estimands that preserve algorithmic regularization. Our proposed fixed-Jacobian bootstrap propagates perturbations along the observed trajectory without repeatedly running the original nonlinear algorithm. We establish uniform first-order expansions and bootstrap validity over possibly infinite sets of contrasts and iterations under stability, entropy and moment conditions. Simulations in sparse linear regression, paired Fisher discrimination, and temporal-difference learning demonstrate accurate simultaneous coverage. A micro-randomized trial application illustrates inference for a treatment trajectory with participant-level clustering.
ER -