Mercurial > hg > smallbox
changeset 210:f12a476a4977 luisf_dev
Added help comments to SMALL_two_step_DL.m
| author | bmailhe |
|---|---|
| date | Wed, 21 Mar 2012 17:25:40 +0000 |
| parents | dfa795944aae |
| children | 0c7c20f3246c 1d134a1b6f95 |
| files | DL/two-step DL/SMALL_two_step_DL.m |
| diffstat | 1 files changed, 14 insertions(+), 2 deletions(-) [+] |
line wrap: on
line diff
--- a/DL/two-step DL/SMALL_two_step_DL.m Wed Mar 21 16:21:18 2012 +0000 +++ b/DL/two-step DL/SMALL_two_step_DL.m Wed Mar 21 17:25:40 2012 +0000 @@ -1,4 +1,16 @@ function DL=SMALL_two_step_DL(Problem, DL) + + %% DL=SMALL_two_step_DL(Problem, DL) learn a dictionary using two_step_DL + % The specific parameters of the DL structure are: + % -name: can be either 'ols', 'opt', 'MOD', KSVD' or 'LGD'. + % -param.learningRate: a step size used by 'ols' and 'opt'. Default: 0.1 + % for 'ols', 1 for 'opt'. + % -param.flow: can be either 'sequential' or 'parallel'. De fault: + % 'sequential'. Not used by MOD. + % -param.coherence: a real number between 0 and 1. If present, then + % a low-coherence constraint is added to the learning. + % + % See dico_update.m for more details. % determine which solver is used for sparse representation % @@ -57,8 +69,8 @@ end % learningRate. If the type is 'ols', it is the descent step of -% the gradient (typical choice: 0.1). If the type is 'mailhe', the -% descent step is the optimal step*rho (typical choice: 1, although 2 works +% the gradient (default: 0.1). If the type is 'mailhe', the +% descent step is the optimal step*rho (default: 1, although 2 works % better). Not used for MOD and KSVD. if isfield(DL.param,'learningRate')