#MLNotes: Confusion Matrix and Model Evaluation Matrices
The subject will be really confusing if you don't understand this. I am still building knowledge, hence expect updates and refinement over time. These are very short and basic notes and if you are an experience reader looking at my article then feel free to leave comments on how I can expand this article to make it better for beginners like me :)
- Accuracy — It is the correctly predicted labels to the total number of labels.
2. Sensitivity — it is the total number of correctly predicted “Yes-es” to the actual number of “Yes-es”
3. Specificity — it is total number of “Noes” correctly predicted to the total number of “No-es”
4. Precision — it is the probability of detecting a ‘‘Yes’’when it is actually a “Yes”.
5. Recall — it is the probability of detecting a ‘Yes’ correctly.
Please note that Precision-Recall is preferred by some businesses and some will prefer Sensitivity-Specificity Matrix.
