class static
analytics~metrics.ClassificationScore

Source: analyticsdoc.js:3431

Class implements several classification measures (precision, recall, F1, accuracy).

Property

scores

Method

push(correct, predicted)

new ClassificationScore(yTrue, yPred)

For evaluating provided categories from binary? classifiers.

Parameters

Name Type Optional Description

yTrue

(Array of number or module:la.Vector)

 

Ground truth (correct) lable(s).

yPred

(Array of number or module:la.Vector)

 

Predicted (estimated) lable(s).

Property

scores

Returns Object containing different classification measures.

Returns

Object scores - Object with different classification socres.

number scores.count - Count.

number scores.TP - Number of true positives.

number scores.TN - Number of true negative.

number scores.FP - Number of false positives.

number scores.FN - Number of false positives.

number scores.all - Number of all results.

number scores.accuracy - Accuracy score. Formula: (tp + tn) / (tp + fp + fn + tn).

number scores.precision - Precision score. Formula: tp / (tp + fp).

number scores.recall - Recall score. Formula: tp / (tp + fn).

number scores.f1 - F1 score. Formula: 2 * (precision * recall) / (precision + recall).

Method

push(correct, predicted)

Adds prediction to the current statistics. Labels can be either integers. or integer array (when there are zero or more then one lables).

Parameters

Name Type Optional Description

correct

number

 

Correct lable.

predicted

number

 

Predicted lable.