A Software Tool For Sparse Estimation Of A General Class Of High-dimensional GLMs

Abstract

Generalized linear models are the workhorse of many inferential problems. Also in the modern era with high-dimensional settings, such models have been proven to be effective exploratory tools. Most attention has been paid to Gaussian, binomial and Poisson settings, which have efficient computational implementations and where either the dispersion parameter is largely irrelevant or absent. However, general GLMs have dispersion parameters φ that affect the value of the log- likelihood. This in turn, affects the value of various information criteria such as AIC and BIC, and has a considerable impact on the computation and selection of the optimal model.The R-package dglars is one of the standard packages to perform high-dimensional analyses for GLMs. Being based on fundamental likelihood considerations, rather than arbitrary penalization, it naturally extends to the general GLM setting. In this paper, we present an improved predictor-corrector (IPC) algorithm for computing the differential geometric least angle regression (dgLARS) solution curve, proposed in Augugliaro et al. (2013) and Pazira et al. (2018). We describe the implementation of a stable estimator of the dispersion parameter proposed in Pazira et al. (2018) for high-dimensional exponential dispersion models. A simulation study is conducted to test the performance of the proposed methods and algorithms. We illustrate the methods using an example. The described improvements have been implemented in a new version of the R-package dglars.