As discussed in the previous chapter, the basic idea underlying linear regression is that some variables' values can be predicted by the following equation describing a line:

Here, the dependent variable *Y* has a linear relationship with a set of *X* values (that is, *X* values that are all raised to the power of 1). Of course, the various *X* values themselves can be nonlinear functions of other predictor variables; thus, by performing linear regression on nonlinear transformations of predictor variables we will be able to model nonlinear relationships in between variables.

The simplest way to extend the ...

Start Free Trial

No credit card required