Class  Description 

ConstrainedLASSObyLARS 
This class solves the constrained form of LASSO by modified least angle regression (LARS) and
linear interpolation:
\[
\min_w \left \{ \left \ Xw  y \right \_2^2 \right \}\) subject to \( \left \ w \right \_1
\leq t
\]

ConstrainedLASSObyQP 
This class solves the constrained form of LASSO (i.e.\(\min_w \left \{ \left \ Xw  y \right
\_2^2 \right \}\)
subject to \( \left \ w \right \_1 \leq t \)) by transforming it into a single quadratic
programming problem with (2 * m + 1) constraints, where m is the number of
columns of the design matrix.

ConstrainedLASSOProblem 
A LASSO (least absolute shrinkage and selection operator) problem focuses on solving an RSS
(residual sum of squared errors) problem with L1 regularization.

UnconstrainedLASSObyCoordinateDescent 
This class solves the unconstrained form of LASSO, that is,
\[
\min_w \left \{ \left \ Xw  y \right \_2^2 + \lambda * \left \ w
\right \_1 \right \}
\]
by Coordinate Descent method.

UnconstrainedLASSObyQP 
This class solves the unconstrained form of LASSO
(i.e.

UnconstrainedLASSOProblem 
A LASSO (least absolute shrinkage and selection operator) problem focuses on solving an RSS
(residual sum of squared errors) problem with L1 regularization.

Copyright © 20102016 Numerical Method Incorporation Limited. All Rights Reserved.