proxGrad: Proximal gradient algorithm

In bozenne/lava.penalty: Penalized Latent Variable Models

Description

Estimate parameters using a proximal gradient algorithm

Usage

proxGrad(start, proxOperator, hessian, gradient, objective,
  control = lava.options()$proxGrad)

Arguments

start	initial values for the parameters
proxOperator	proximal operator corresponding to the penalization applied to the log likelihood
hessian	second derivative of the likelihood given by lava. Only used to estimate the step parameter of the algorithm when step = NULL
gradient	first derivative of the likelihood given by lava.
objective	likelihood given by lava. Used to adjust the step parameter when using backtracking
control	settings for the proximal gradient algorithm. See lava.options.

References

Bech and Teboulle - 2009 A Fast Iterative Shrinkage-Thresholding Algorithm Li 2015 - Accelerated Proximal Gradient Methods for Nonconvex Programming Simon 2013 - A sparse group Lasso

bozenne/lava.penalty documentation built on May 13, 2019, 1:41 a.m.