freediscovery.categorization.Categorizer

class freediscovery.categorization.Categorizer(cache_dir=u'/tmp/', dsid=None, mid=None, cv_scoring=u'roc_auc', cv_n_folds=3)

Document categorization model

The option use_hashing=True must be set for the feature extraction. Recommended options also include, use_idf=1, sublinear_tf=0, binary=0.

Parameters:

• cache_dir (str) – folder where the model will be saved
• dsid (str, optional) – dataset id
• mid (str, optional) – model id
• cv_scoring (str, optional, default='roc_auc') – score that is used for Cross Validation, cf. sklearn
• cv_n_folds (str, optional) – number of K-folds used for Cross Validation

__init__(cache_dir=u'/tmp/', dsid=None, mid=None, cv_scoring=u'roc_auc', cv_n_folds=3)

Methods

__init__([cache_dir, dsid, mid, cv_scoring, ...])
delete()	Delete a trained model
get_dsid(cache_dir, mid)
get_params()	Get model parameters
get_path(mid)
list_models()
predict([chunk_size])	Predict the relevance using a previously trained model
train(index, y[, method, cv])	Train the categorization model

delete()

Delete a trained model

get_params()

Get model parameters

predict(chunk_size=5000)

Predict the relevance using a previously trained model

Parameters:chunck_size (int) – chunck size

train(index, y, method=u'LinearSVC', cv=None)

Train the categorization model

Parameters:

• index (array-like, shape (n_samples)) – document indices of the training set
• y (array-like, shape (n_samples)) – target binary class relative to index
• method (str) – the ML algorithm to use (one of “LogisticRegression”, “LinearSVC”, ‘xgboost’)
• cv (str) – use cross-validation

Returns:

• cmod (sklearn.BaseEstimator) – the scikit learn classifier object
• Y_train (array-like, shape (n_samples)) – training predictions