en_algorithm_options: Control the Algorithm to Compute (Weighted) Least-Squares...

In pense: Penalized Elastic Net S/MM-Estimator of Regression

Control the Algorithm to Compute (Weighted) Least-Squares Elastic Net Estimates

Description

The package supports different algorithms to compute the EN estimate for weighted LS loss functions. Each algorithm has certain characteristics that make it useful for some problems. To select a specific algorithm and adjust the options, use any of the en_***_options functions.

Details

• en_lars_options(): Use the tuning-free LARS algorithm. This computes exact (up to numerical errors) solutions to the EN-LS problem. It is not iterative and therefore can not benefit from approximate solutions, but in turn guarantees that a solution will be found.

• en_cd_options(): Use an iterative coordinate descent algorithm which needs O(n p) operations per iteration and converges sub-linearly.

• en_admm_options(): Use an iterative ADMM-type algorithm which needs O(n p) operations per iteration and converges sub-linearly.

• en_dal_options(): Use the iterative Dual Augmented Lagrangian (DAL) method. DAL needs O(n^3 p^2) operations per iteration, but converges exponentially.

pense documentation built on Feb. 16, 2023, 9:36 p.m.