r-qut

Thresholding based tests for null hypothesis of the form A beta =c, and the Quantile Universal Threshold (QUT) for lasso regularization of Generalized Linear Models (GLM) and square-root lasso to obtain a sparse model with a good compromise between high true positive rate and low false discovery rate. Giacobino et al. (2017) <doi:10.1214/17-EJS1366>. Sardy et al. (2017) <arXiv:1708.02908>.