Choosing the type of regression model

With all of these options, how do you choose the right type of regression model?

The type of regression model you choose depends on what your expected output is. If you are looking for just a Boolean (that is, 0 or 1) value, logistic regression models should be used like in the file classification application we will be writing later in this chapter. In addition, if you are looking to return a specific pre-defined range of values, perhaps a car type such as coupe, convertible, or hatchback, a logistic regression model is the correct model to choose from. 

Conversely, linear regression models return a numeric value, such as the employment duration example we will explore later in this chapter. 

So, to summarize, we have the following:

• If your output is a Boolean value, use a logistic regression model.
• If your output is comprised of a preset range type of values (akin to an enumeration), use a logistic regression model.
• If your output is a numeric unknown value, use a linear regression model.