ABSTRACT
This chapter covers a different type of modeling framework, known as logistic regression, and describes some of the basic concepts associated with logistic regression, such as the relationship of probability and odds, and the concept of an odds ratio. Logistic regression models are appropriate when modeling cases with binary type outcomes. Unlike the simple linear regression model, there is no simple formula for fitting the logistic regression model. A method known as maximum likelihood estimation is used for estimating the parameters of the logistic regression model. The approach takes advantage of the theoretical properties of the logistic regression model to create something called a likelihood function. The logistic regression model is the most commonly used model for modeling outcomes with only two values. It is very popular in modeling fraud. There are many popular software packages that can fit the logistic regression model and many opportunities within loss prevention to use the model.
