lsgl-package: Linear Multiple Output Using Sparse Group Lasso.

Description Details Author(s) Examples


Simultaneous feature selection and parameter estimation for linear multiple output. The algorithm finds the sparse group lasso penalized maximum likelihood estimator. This result in feature and parameter selection, and parameter estimation. Use of parallel computing for cross validation and subsampling is supported through the 'foreach' and 'doParallel' packages.


The sparse gorup lasso linear mltiple output estimator is defined as

\frac{1}{N}\|Y-Xβ\|_F^2 + λ ≤ft( (1-α) ∑_{J=1}^m γ_J \|β^{(J)}\|_2 + α ∑_{i=1}^{n} ξ_i |β_i| \right)

where \|\cdot\|_F is the frobenius norm. The vector β^{(J)} denotes the parameters associated with the J'th group of features. The group weights are denoted by γ \in [0,∞)^m and the parameter weights by ξ \in [0,∞)^n.


Martin Vincent



lsgl documentation built on May 29, 2017, 11:43 a.m.