Add classification and regression metrics
msluszniak opened this issue · comments
I want to add all the existing metrics which are present in Sci-kit Learn.
Classification Metrics
- Accuracy
- AUC
- Average Precision (AP)
- Balanced Accuracy
- Brier Score Loss
- Cohen Kappa Score
- Confusion Matrix
- Discounted Cumulative Gain (DCG)
- Detection Error Tradeoff Curve (DET curve)
- F1 Score
- F-beta Score
- Hamming Loss
- Log Loss
- Matthews correlation coefficient (MCC)
- Normalized Discounted Cumulative Gain (Normalized DCG)
- Precision Recall Curve
- Precision Recall Fscore Support
- Precision
- Recall
- ROC-AUC
- ROC Curve
- Top-k Accuracy Score
- Zero-one Classification Loss
Regression Metrics
- Explained Variance Score
- Maximum Residual Error
- Mean Absolute Error (MAE)
- Mean Squared Error (MSE)
- Mean Squared Log Error
- Mean Absolute Percentage Error (MAPE)
-
$R^2$ Score - Mean Poisson deviance
- Mean Gamma deviance
- Mean Tweedie deviance
-
$D^2$ - Fraction of Tweedie Deviance Explained - Mean Pinball Loss
-
$D^2$ - Fraction of Pinball Loss Explained -
$D^2$ - Fraction of Absolute Error Explained
We can do the same for distances after (look #149).
I'd like to contribute to this project. I will start working on a PR for "Discounted Cumulative Gain (DCG)"
I'd like to contribute to this project. I will start working on a PR for "Discounted Cumulative Gain (DCG)"
Really cool! Feel free to go :)
I have Normalized DCG in progress.
I have mean pinball loss in progress.
We added all metrics, thank all of you who contributed ❤️