preview

Classi Cation Performance Evaluation Report

Satisfactory Essays
To evaluate our systems classi cation performance, i.e. the proportion of correctly classi ed tweets for a given test set, we used four common information retrieval (IR) evaluation measures, including recall (R), Precision (P), F1 Score, and Accuracy [73]. To explain these evaluation measures, we use a confusion matrix depicted in gure 5.1. 1 Classified Positive Classified Negative Actual Positive TP FN Actual Negative FP TN Figure 5.1: Confusion Matrix Where, TP (True Positive) ) Number of correct classi cations of the positive samples FN (False Negative) ) Number of incorrect classi cations of positive samples FP (False Positive) ) Number of incorrect classi cations of negative samples TN (True Negative) ) Number of correct classi cations…show more content…
In another way, it is the number of positive predictions divided by the total number of positive class values predicted. It is also called the Positive Predictive Value (PPV). Precision can be thought of as a measure of a classi ers exactness. A low precision can also indicate a large number of False Positives. Precision, P = TP TP + FP Recall: Recall is the number of True Positives divided by the number of True Positives and the number of False Negatives. In another way it is the number of positive 48 5.3 Results with Supervised Feature Selection predictions divided by the number of positive class values in the test data. It is also called Sensitivity or the True Positive Rate. Recall can be thought of as a measure of a classi ers completeness. A low recall indicates many False Negatives. Recall, R = TP TP + FN F1 Score: The F1 score, also called the F Score or the F Measure, conveys the balance between the precision and the recall. The traditional F1 score (F-measure or balanced F-score) is the harmonic mean of precision and recall: F1 Score, = 2  P  R P + R Accuracy: The accuracy is the proportion of true results (both true positives and true negatives) among the total number of cases examined. Accuracy simply
Get Access