Reproduce “Finite-Sample Distribution Theory and Efficient Large-Scale Inference for Online Quantile Regression” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ziyang Wei, Jiaqi Li, Lan Wang, Wei Biao Wu
The authors develop a single-pass inference method for streaming quantile regression. It updates estimates with constant-learning-rate stochastic gradient descent, discards early iterates through suffix averaging, and uses variation across independent runs to form confidence intervals without covariance estimation or bootstrap resampling. At n = 10⁷, parallel inference took 27.660 seconds versus 89.000 for random scaling. Its guarantees depend on stated assumptions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.