Data Mining

2

Linear and Nonlinear Regression Models

Regression models capture how one or more target variables vary with one or more attribute variables. They can be used to predict the values of the target variables using the values of the attribute variables. In this chapter, we introduce linear and nonlinear regression models. This chapter also describes the least-squares method and the maximum likelihood method of estimating parameters in regression models. A list of software packages that support building regression models is provided.

2.1    Linear Regression Models

A simple linear regression model, as shown next, has one target variable y and one attribute variable x:

$y_i={\text{β}}_0+{\text{β}}_1{x}_i+{\text{ε}}_i$

(2.1)

where

(x_i, y_i) denotes the ith observation of ...