Explain how you would choose between a logistic regression model and a random forest for a classification problem with highly imbalanced data.
