Smoothed LASSO Regression via Nesterov Smoothing

[Up] [Top]

Documentation for package ‘smoothedLasso’ version 1.3

Help Pages

objFunction Auxiliary function defining the LASSO objective function.
objFunctionGradient Auxiliary function which computes the (non-smooth) gradient of the LASSO objective function with respect to beta.
objFunctionSmooth Auxiliary function defining the smoothed LASSO objective function.
objFunctionSmoothGradient Auxiliary function which computes the gradient of the smoothed LASSO objective function with respect to beta.
solveSmoothedLASSO Minimize the smoothed LASSO objective function with respect to beta using BFGS.
solveSmoothedLASSOSequence Minimize the smoothed LASSO objective function with respect to beta using the progressive smoothing algorithm.
solveUnsmoothedLASSO Minimize the unsmoothed LASSO objective function with respect to beta. Three options are available: BFGS with analytical gradient (method=0), BFGS with numerical gradient (method=1), and simulated annealing which is gradient free (method=2). The default is method=0.