probml / pmtk3

Probabilistic Modeling Toolkit for Matlab/Octave.
MIT License
1.55k stars 797 forks source link

Some question in the document:pmtk3/toolbox/SupervisedModels/logreg/logregFit.m #140

Open sxjaxs opened 5 years ago

sxjaxs commented 5 years ago

In the line 12 the author defines the three default values of regType:“ % regType ... 'L1' or 'L2' or 'none'”,but when solving optimization problems from the line 133 to the line 141,there are only the item 'L1' and 'L2',how can I got the posterior distribution of the parameter W without the regularization terms?Has anyone studied this question? thanks a lot!!!