I create a balanced data set and use it in cyber bullying detection ,All the performance metrics ( ACC, F1, Recall and Precision) are the same what is the interpretation ??
If they are the same, there are two possibilities, (a) dataset is not consistent so you have to reimplement your algorithm on another dataset or on a portion of the dataset, (b) if matrix table i.e., TP, TN,... are different then it is occurred by chance so all mertices have the same impact on performance, in this case you can interpret according to the resulting value,
Please read the following
https://blog.exsilio.com/all/accuracy-precision-recall-f1-score-interpretation-of-performance-measures/#:~:text=Recall%20(Sensitivity)%20-%20Recall%20is,observations%20in%20actual%20class%20-%20yes.&text=F1%20score%20-%20F1%20Score%20is,and%20false%20negatives%20into%20account.
Best wishes
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers