API reference

class mord.LogisticAT(alpha=1.0, verbose=0, maxiter=10000)

Classifier that implements the ordinal logistic model (All-Threshold variant)

Parameters:

alpha: float :

Regularization parameter. Zero is no regularization, higher values increate the squared l2 regularization.

References

J. D. M. Rennie and N. Srebro, “Loss Functions for Preference Levels : Regression with Discrete Ordered Labels,” in Proceedings of the IJCAI Multidisciplinary Workshop on Advances in Preference Handling, 2005.

Methods

class mord.LogisticIT(alpha=1.0, verbose=0, maxiter=10000)

Classifier that implements the ordinal logistic model (Immediate-Threshold variant)

The score is the same as in multiclass classification methods (i.e. metrics.accuracy_score)

Parameters:

alpha: float :

Regularization parameter. Zero is no regularization, higher values increate the squared l2 regularization.

References

J. D. M. Rennie and N. Srebro, “Loss Functions for Preference Levels : Regression with Discrete Ordered Labels,” in Proceedings of the IJCAI Multidisciplinary Workshop on Advances in Preference Handling, 2005.

Methods

class mord.LAD(epsilon=0.0, tol=0.0001, C=1.0, loss='l1', fit_intercept=True, intercept_scaling=1.0, dual=True, verbose=0, random_state=None, max_iter=1000)

Least Absolute Deviation

Methods

class mord.MulticlassLogistic(alpha=1.0, verbose=0, maxiter=10000)

Methods

class mord.OrdinalRidge(alpha=1.0, fit_intercept=True, normalize=False, copy_X=True, max_iter=None, tol=0.001, solver='auto')

Overwrite Ridge from scikit-learn to use the (minus) absolute error as score function.

(see https://github.com/scikit-learn/scikit-learn/issues/3848 on why this cannot be accomplished using a GridSearchCV object)

Methods