Constructor
new LogisticRegressionClassifier()
Methods
classifyObservation(srcObservation, classifications)
Given an observation and an array for inserting the results,
it calculates the score of the observation for each of the classifications
and fills the array with the result objects.
Parameters:
Name | Type | Description |
---|---|---|
srcObservation |
Object | Source observation. |
classifications |
Array.<Object> | Array of classifications. |
newClassification(observation, indexClassification)
Given an observation vector and the index of one of the classifications,
it returns an object that contains the label of the classification and
the score of the vector for this classification.
Parameters:
Name | Type | Description |
---|---|---|
observation |
Vector | Observation vector. |
indexClassification |
Number | Index of the classification. |
(async) train()
Train the logistic regression clasifier, that means
that it calculates the thetas that relates all the features
with the classifications, so when a new vector of features
is the input to classify, these thetas are the weights for the
calculation of the scores of each classification.