Class: Classifier

Classifier(settings)

Class for a generic classifier. This is an abstract class that must be implemented by subclasses that contains the real classifier algorithm.

Constructor

new Classifier(settings)

Constructor of the class. Initialize the basic properties and structure of any classifier.
Parameters:
Name Type Description
settings Object Settings for initializing the instance.
Source:

Methods

addLabel(label) → {Array.<String>}

Adds a new label to the observation tree. If the label already exists, return the existing one without creating it again.
Parameters:
Name Type Description
label String Label to be created or getted.
Source:
Returns:
List of observations assigned to this label.
Type
Array.<String>

addObservation(observation, label)

Adds a new observation to the classifier.
Parameters:
Name Type Description
observation String Observation to be added.
label String Label of the observation.
Source:

classifyObservation()

Classify one observation.
Source:

clear()

Clears the content of the instance. This is done by initializing the observations object, the labels array and the observation count.
Source:

createClassificationMatrix() → {Array.<Array.<Number>>}

Creates a matrix filled with zeros, that relate every single observation with every single label.
Source:
Returns:
A bidimensional array where x is the observation and y is the label, filled to zeros.
Type
Array.<Array.<Number>>

getBestClassification(observation) → {Object}

Given an observation, get the label and score of the best classification.
Parameters:
Name Type Description
observation String Observation to be classified.
Source:
Returns:
Best classification of the observation.
Type
Object

getClassifications(observation) → {Array.<Object>}

Get all the labels and score for each label from one observation.
Parameters:
Name Type Description
observation String Observation to be classified.
Source:
Returns:
Sorted array of classifications, that means label and the score.
Type
Array.<Object>

recalculateObservationCount()

Iterate all the observations to calculate the total observation count.
Source:

removeObservation(observation, label)

Removes an observation. The label of the observation can be passed or can be omitted. When omitted, it loops over all labels tryin to remove the given observation.
Parameters:
Name Type Description
observation String Observation to be removed.
label String Label of the observation, or undefined to iterate over all labels.
Source:

removeObservationByLabel(observation, label)

Removes an observation from the observation list of a label.
Parameters:
Name Type Description
observation String Observation to be removed.
label String Label where we want the observation to be removed.
Source: