Imprecise Logistic Regression

Imprecise Logistic Regression class mimicing scikit-learn's LogisticRegression class

---

Usage

The usage is identical to scikit-learn except you have to tell the class where the intervals are

from ImpLogReg import ImpLogReg
ilr = ImpLogReg(uncertain_class = True)
ilr.fit(data,results)

---

Documentation

class ImpLogReg(uncertain_data = False, uncertain_class = False, **kwargs)

Parameters

uncertain_data : bool, default = False

Whether the imprecise model has interval datapoints.

uncertain_class: bool, default = False

Whether the imprecise model has uncertain results.

**kwargs

kwargs passed to sklearn.linear_model.LogisticRegression() function.

Attributes

models: dict

Dictionary containing the set of LR models trained on the dataset.

coef_: list

Set of possible coefficients of the features in the decision functions.

intercept_: list

Set of coefficient of the features in the decision functions.

n_iter_: list

Actual number of interations for all LR models in the set

Methods

decision_function(X)	Predict confidence scores for samples.
densify()	Convert coefficient matrix to dense array format.
fit(X, y, sample_weight, catagorical)	Fit the model according to the given training data.
get_params()	Get parameters for this estimator.
predict(X)	Predict class labels for samples in X.
predict_log_proba(X)	Predict logarithm of probability estimates.
predict_proba(X)	Probability estimates.
score(X, y , sample_weight)	Return the mean accuracy on the given test data and labels.
set_params(**params)	Set the parameters of this estimator.
sparsify()	Convert coefficient matrix to sparse format.

decision_function(X)

Predict confidence scores for samples.

The confidence score for a sample is proportional to the signed distance of that sample to the hyperplane.

Parameters

X: array-like or sparse matrix, shape (n_samples, n_features)

Samples Passed to

Returns

list of pba.Interval objects

Interval of confidence score for each sample for all m models

See scikit-learn.linear_model.LogisticRegression().decision_function()

densify()

Converts coefficient matrix to dense array format for all LR models

Returns

self

See scikit-learn.linear_model.LogisticRegression().densify()

fit(X, y, sample_weight = None, catagorical = [])

Fit the model according to the given training data.

Parameters

X: {array-like, sparse matrix} of shape (n_samples, n_features)

Training vector, where n_samples is the number of samples and n_features is the number of features. If uncertain_data is true then will expect

y: array-like of shape (n_samples,)

Target vector relative to X.

sample_weight: array-like of shape (n_samples,) default=None

Array of weights that are assigned to individual samples. If not provided, then each sample is given unit weight.

catagorical: array-like default = []

Array containing a list of features that contain catagorical data.

Returns

self

Dictionary of fitted estimators

See scikit-learn.linear_model.LogisticRegression().fit()

get_params()

Get parameters for this estimator.

Returns

params: dict

See scikit-learn.linear_model.LogisticRegression().get_params()

predict(X)

Predict class labels for this estimator

Parameters

X: array-like or sparse matrix, shape (n_samples, n_features)

Samples

Returns

list of pba.Logical objects

Predicted class labels per sample. 0 if sample is always 0, 1 if sample is always 1 or pba.Logical(0,1) otherwise

See scikit-learn.linear_model.LogisticRegression().predict()

predict_log_proba(X)

Predict class labels for this estimator

Parameters

X: array-like or sparse matrix, shape (n_samples, n_features)

Samples

Returns

array-like of shape (n_samples, n_classes) containing pba.Interval objects

Returns the Interval log-probability of the sample for each class in the model, where classes are ordered as they are in self.classes_.

See scikit-learn.linear_model.LogisticRegression().predict_log_proba()

predict_proba(X)

Predict class labels for this estimator

Parameters

X: array-like or sparse matrix, shape (n_samples, n_features)

Samples

Returns

array-like of shape (n_samples, n_classes) containing pba.Interval objects

Returns the Interval probability of the sample for each class in the model, where classes are ordered as they are in self.classes_.

See scikit-learn.linear_model.LogisticRegression().predict_proba()

score(X, y , sample_weight)

Return the mean accuracy on the given test data and labels.

In multi-label classification, this is the subset accuracy which is a harsh metric since you require for each sample that each label set be correctly predicted.

Parameters

X: {array-like, sparse matrix} of shape (n_samples, n_features)

Training vector, where n_samples is the number of samples and n_features is the number of features. If uncertain_data is true then will expect

y: array-like of shape (n_samples,)

Target vector relative to X.

sample_weight: array-like of shape (n_samples,) default=None

Array of weights that are assigned to individual samples. If not provided, then each sample is given unit weight.

Returns

score: pba.Interval

Interval containing the minimum and maximum accuracy score for all models

See scikit-learn.linear_model.LogisticRegression().score()

set_params(**params)

Parameters

params: dict

Estimator parameters.

Returns

self: Estimators.

See scikit-learn.linear_model.LogisticRegression().set_params()

sparsify()

Convert coefficient matrix to sparse format for each model

Returns

self: Fitted estimators.

See scikit-learn.linear_model.LogisticRegression().sparsify()