Table Of Contents

Previous topic

mvpa.clfs.model_selector

Next topic

mvpa.clfs.ridge

This Page

Quick search

mvpa.clfs.plr

Penalized logistic regression classifier.

The comprehensive API documentation for this module, including all technical details, is available in the Epydoc-generated API reference for mvpa.clfs.plr (for developers).

PLR

class mvpa.clfs.plr.PLR(lm=1, criterion=1, reduced=False, maxiter=20, **kwargs)

Bases: mvpa.clfs.base.Classifier

Penalized logistic regression Classifier.

Initialize a penalized logistic regression analysis

Parameters:
  • lm (int) – the penalty term lambda.
  • criterion (int) – the criterion applied to judge convergence.
  • reduced (Bool) – if not False, the rank of the data is reduced before performing the calculations. In that case, reduce is taken as the fraction of the first singular value, at which a dimension is not considered significant anymore. A reasonable criterion is reduced=0.01
  • maxiter (int) – maximum number of iterations. If no convergence occurs after this number of iterations, an exception is raised.

See also

Derived classes might provide additional methods via their base classes. Please refer to the list of base classes (if it exists) at the begining of the PLR documentation.

Full API documentation of PLR in module mvpa.clfs.plr.