Performance Evaluation Metrics

Python 3

Classification Performance Metrics

Regression Performance Metrics

Theory

For Classification???

Confusion Matrix

		Actual value
		Positive	Negative
Predicted	Positive	True Positive	False Positive
Value	Negative	False Negative	True Negative

Accuracy

Precision

Precision is used when the cost of a false positive is higher than the cost of a false negative.

Recall

Recall is used when the cost of a false negative is higher than the cost false positive.

F1-Score

The F1-Score is an optimal blend of precision and recall.

For Regression???

Correlation Coefficient (r)

R

Mean Absolute Error (MAE)

Root Mean Square Error (RMSE)

Willmott’s Index of Agreement (WI)

Nash-Sutcliffe coefficient (ENS)

Legates and Mc-Cabes Index (ELM)

Reference

https://scikit-learn.org/stable/modules/generated/sklearn.metrics.confusion_matrix.html

https://www.kdnuggets.com/2020/04/performance-evaluation-metrics-classification.html