List of metrics First Tier:
- False Positive - Type I
- False Negative - Type II
- True Positive
- True Negative
Second Tier: Primary issues with 2nd tier - not easy to remember and have various names for the same thing: is there a way to prevent this, i.e. create some type of way to help people understand the concept so well that they don't necessarily need to use the specific term - i.e. they will know what it does/what it tells us (which is more important)
- Accuracy - (TP + TN)/N
- Precision - TP/(TP + FP) - AKA Positive Predictive Value
- Specificity
- TN/(all negative samples (not predicted))
- AKA True negative Rate
- Sensitivity
- AKA: Recall, True Positive Rate
- Formula:
- False negative rate
- Miss Rate
- False positive rate
- Fall Out
Third Tier:
- ROC/AUC
- Positive Likelihood Ratio
- Negative Likelihood Ratio
- Diagnostic Odds Ratio