ABSTRACT
Linear regression modeling is well suited to predicting continuous data where the outcome y is a real number (i.e., y ∈ ℝ). Logistic regression is a modeling technique for binary outcomes (i.e., yes/no, true/false, 1/0). Such outcomes are needed in many domains: public health officials might want to know the likelihood that a person will contract COVID-19 if she is a doctor in Ontario;a hospital would like to know if a discharged patient is more likely to be readmitted or not;a company would like to know if a customer visiting its website is more likely to order;a bank would like to know if a customer is more likely to default on a loan or not. Logistic regression has been much used in the medical field and yielded impressive results [1–10]. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.