Joint Feature Screening via Sparse MLE
Description
Variable selection techniques are essential tools for model selection and estimation in high-dimensional statistical models. Sparse Maximal Likelihood Estimator (SMLE) (Xu and Chen (2014)) provides an efficient implementation for the joint feature screening method on high-dimensional generalized linear models. It also conducts a post-screening selection based on a user-specified selection criterion. The algorithm uses iterative hard thresholding along with parallel computing.