7
2.7 Evaluation meters
To evaluate the efficacy of the models, the confusion matrix and the
classification report were used. The confusion matrix is a compact
representation of the results of a classification task prediction. It is
used to evaluate the overall performance in terms of accuracy,
precision, recall, and F1-score. A combination of correct (True) and
incorrect (False) classifications is used to determine these measures.
In this context, the projected classes may be generally referred to as
”Positive” or ”Negative.” Total four parameters are used, True Positive
(TP) (case is positive, prediction is also positive), True Negative (TN)
(case is negative and prediction is also negative), False Positive (FP)
(case is negative and prediction is positive) and False Negative (FN)
(case is positive and false negative). Using these four parameters
evaluation meters, namely, accuracy, precision, recall and F1-score,
were computed.