9.4 Chapter 4

AIC/BIC: The AIC and BIC are information criteria that can be used to compare models, where we want a model with a lower AIC or BIC.

False Positive Rate: The relative frequency of predicting the event would happen, given that the event didn’t happen (count of false positives/count of times event did not actually happen).

False Negative Rate: The relative frequency of predicting the event wouldn’t happen, given that the event did happen (count of false negatives/count of times event did actually happen).

Logistic Regression: A model that is used to predict a binary categorical variable (only has two outcomes).

Odds: The chance of an event divided by the chance the event does not happen.

Odds Ratio: A ratio of odds an event will happen, comparing between two subgroups.

Sensitivity: The relative frequency of predicting the event would happen, given that the event did happen (count of true positives/count of times event did actually happen).

Specificity: The relative frequency of predicting the event would not happen, given that the event did not happen (count of true negatives/count of times event did not actually happen).

Slope Interpretation in Logistic Regression Model: If there are no interaction terms: For a 1 unit increase in X, we’d expect the estimated odds of an event to change by a multiplicative factor of \(e^{b_j}\), keeping all other variables fixed. If there are interaction terms: write out the model for subgroups to determine interpretation.