ML6 - Model Assessment and Deployment#
6.1 - Model Assessment and Comparison#
Prediction Type |
Assessment Measures |
|---|---|
Decision |
Accuracy, Misclassification, Profit/loss |
Ranking |
C-statistic, Gini coefficient |
Estimate |
Average square error, Root mean square error, Schwarz criterion (SC), Akaike’s information criterion (AIC) |
\[ \text{Sensitivity} = \frac{\text{TP}}{\text{TP} + \text{FN}} \]
\[ \text{Specificity} = \frac{\text{TN}}{\text{TN} + \text{FP}} \]