Evaluating a logistic regression model
There are multiple metrics we can use to evaluate a logistic regression model. These are the metrics we use to determine the goodness of fit (over the test set), which needs to be differentiated from the CEL we use to train the model (over the training set).
The following list provides the commonly used metrics:
- Accuracy rate: This is the proportion of the number of correctly predicted observations made by the model out of the count of all observations. Since a correct prediction can be either a true positive or a true negative, the accuracy is calculated by summing up the true positives and true negatives and dividing the total number of observations.
- Error rate: This is the proportion of incorrectly predicted observations made by the model over the total observations. An incorrect prediction can be a false positive or a false negative. It is calculated as 1 - accuracy rate; that is, the error rate is the complement of the accuracy...
