Implementation in Python of the conformalized quantile regression (CQR) method for constructing marginal, distribusion-free prediction intervals
Python and R scripts for sampling exact knockoffs for graphical models via the method of Metropolized knockoff sampling
R package with efficient algorithms to generate knockoffs for hidden Markov models, with special support for genetic data
R and MATLAB packages implementing the knockoff filter, a novel model selection strategy for variable selection which provably controls Type-I errors such as the FDR (false discovery rate)
R and MATLAB packages implementing SLOPE, or Sorted L-One Penalized Estimation. The sorted L1 norm is useful for statistical estimation and testing, particularly for variable selection in the linear model.
Templates for First-Order Conic Solvers (TFOCS, pronounced tee-fox) is a MATLAB package that provides a set of templates, or building blocks, that can be used to construct efficient, customized solvers for a variety of models.
Singular Value Thresholding (SVT) is an algorithm to minimize the nuclear norm of a matrix, subject to certain types of constraints.