Abstract
By introducing a weighted score function, we propose a new penalized method, similar to square root lasso, to study sparse Poisson regression problems. The corresponding new estimator not only has $\ell _{1}$ consistency but also enjoys the tuning free property. We further verify our theoretical results by numerical simulations and apply them to an image reconstruction problem.
Citation
Jinzhu Jia. Fang Xie. Lihu Xu. "Sparse Poisson regression with penalized weighted score function." Electron. J. Statist. 13 (2) 2898 - 2920, 2019. https://doi.org/10.1214/19-EJS1580
Information