pydl85.supervised.classifiers.Boosting_Model

class pydl85.supervised.classifiers.Boosting_Model(value)[source]

The mathematical model solved by the boosting algorithm

Variables:

  • MODEL_LP_RATSCH – Model used in Ratsch paper. The regulator value is between ]0; 1]

  • MODEL_LP_DEMIRIZ – Model used in Demiriz paper. The regulator value is between ]1/n_instances; +infty]

  • MODEL_LP_AGLIN – Model proposed by Aglin. The regulator value is between [0; 1]

  • MODEL_QP_MDBOOST – regulator used in MDBOOST paper. The regulator value is between ]1/n_instances; +infty]