Reproduce “A data-dependent weighted LASSO under Poisson noise” 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
Xin Jiang, Patricia Reynaud-Bouret, Vincent Rivoirard, Laure Sansonnet, Rebecca Willett
The authors develop a weighted LASSO estimator for sparse inverse problems with Poisson observations. It transforms measurements and uses concentration bounds to estimate coordinate-specific penalty weights, while rescaling the design keeps the problem compatible with standard LASSO solvers. In a simplified genomic convolution simulation, weighted LASSO has MSE scaling proportional to 1/√p, versus log²(p)/√p for standard LASSO, under specified conditions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.