| Metric | Purpose | Used for | Range/Scale | Example |
|---|---|---|---|---|
| Accuracy | Overall correctness of predictions | Classification | [0, 1] | 0.85 (85%) |
| Precision | True positives among predicted positives | Binary/Multiclass Classification | [0, 1] | 0.75 (75%) |
| Recall (Sensitivity) | True positives among actual positives | Binary/Multiclass Classification | [0, 1] | 0.90 (90%) |
| F1 Score | Harmonic mean of precision and recall | Binary/Multiclass Classification | [0, 1] | 0.82 |
| ROC-AUC | Area under the Receiver Operating Characteristic curve | Binary Classification | [0, 1] | 0.89 |
| Mean Absolute Error (MAE) | Average absolute difference between actual and predicted values | Regression | [0, ∞) | 5.2 |
| Mean Squared Error (MSE) | Average squared difference between actual and predicted values | Regression | [0, ∞) | 42.3 |
| R-squared (R2) | Proportion of variance explained by the model | Regression | [-∞, 1] | 0.75 |
| Confusion Matrix | True positives, true negatives, false positives, false negatives | Binary/Multiclass Classification | Integers | See matrix example |
| Log Loss (Cross-Entropy) | Evaluation metric for probabilistic predictions | Binary/Multiclass Classification | [-∞, ∞) | 0.45 |
| Mean Average Precision (MAP) | Average precision across multiple classes or queries | Information Retrieval | [0, 1] | 0.78 |
About The Author
DrJha