mindmeld.models.text_models module

This module contains all code required to perform multinomial classification of text.

class mindmeld.models.text_models.PytorchTextModel(config)

Bases: mindmeld.models.model.PytorchModel

evaluate(examples, labels)

Evaluates a model against the given examples and labels

Parameters:

• examples -- A list of examples to predict
• labels -- A list of expected labels

Returns:

an object containing information about the evaluation

Return type:

ModelEvaluation

fit(examples, labels, params=None)

classmethod load(path)

predict(examples, dynamic_resource=None)

predict_proba(examples, dynamic_resource=None)

ALLOWED_CLASSIFIER_TYPES = ['embedder', 'cnn', 'lstm']

class mindmeld.models.text_models.TextModel(config)

Bases: mindmeld.models.model.Model

evaluate(examples, labels)

Evaluates a model against the given examples and labels

Parameters:

• examples -- A list of examples to predict
• labels -- A list of expected labels

Returns:

an object containing information about the evaluation

Return type:

ModelEvaluation

fit(examples, labels, params=None)

Trains this model.

This method inspects instance attributes to determine the classifier object and cross-validation strategy, and then fits the model to the training examples passed in.

Parameters:

• examples (ProcessedQueryList.*Iterator) -- A list of examples.
• labels (ProcessedQueryList.*Iterator) -- A parallel list to examples. The gold labels for each example.
• params (dict, optional) -- Parameters to use when training. Parameter selection will be bypassed if this is provided

Returns:

Returns self to match classifier scikit-learn interfaces.

Return type:

(TextModel)

get_feature_matrix(examples, y=None, fit=False, dynamic_resource=None)

Transforms a list of examples into a feature matrix.

Parameters:examples (list) -- The examples. Returns:tuple containing:

• (numpy.matrix): The feature matrix.
• (numpy.array): The group labels for examples.

Return type:(tuple)

inspect(example, gold_label=None, dynamic_resource=None)

This class takes an example and returns a 2D list for every feature with feature

name, feature value, feature weight and their product for the predicted label. If gold label is passed in, we will also include the feature value and weight for the gold label and returns the log probability of the difference.

Parameters:

• example (Query) -- The query to be predicted
• gold_label (str) -- The gold label for this string
• dynamic_resource (dict, optional) -- A dynamic resource to aid NLP inference

Returns:

A 2D array that includes every feature, their value, weight and probability

Return type:

(list of lists)

classmethod load(path)

predict(examples, dynamic_resource=None)

predict_log_proba(examples, dynamic_resource=None)

predict_proba(examples, dynamic_resource=None)

select_params(examples, labels, selection_settings=None)

Selects the best set of hyper-parameters for a given set of examples and true labels

through cross-validation

Parameters:

• examples -- A list of example queries
• labels -- A list of labels associated with the queries
• selection_settings -- A dictionary of parameter lists to select from

Returns:

A dictionary of optimized parameters to use

Return type:

dict

view_extracted_features(example, dynamic_resource=None)

ACCURACY_SCORING = 'accuracy'

ALLOWED_CLASSIFIER_TYPES = ['logreg', 'dtree', 'rforest', 'svm']

DECISION_TREE_TYPE = 'dtree'

LOG_REG_TYPE = 'logreg'

RANDOM_FOREST_TYPE = 'rforest'

SVM_TYPE = 'svm'

class mindmeld.models.text_models.TextModelFactory

Bases: mindmeld.models.model.AbstractModelFactory

static get_model_cls(config: mindmeld.models.model.ModelConfig)