You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Added support of sparse matrix x of class sparseMatrix (provided by the {Matrix} package) for abclass() and predict.abclass().
Added new functions named cv.abclass() and et.abclass() for training and tuning the angle-based classifiers with cross-validation and an efficient tuning procedure for lasso-type algorithms, respectively. See the corresponding function documentation for details.
Added experimental classifiers with sup-norm penalties. See the functions supclass() and cv.supclass() for details.
Major Changes
Simplified the function abclass() and moved the tuning procedure by cross-validation to the function cv.abclass().
Minor Changes
Changed the default values of the following arguments for abclass.control().