ABSTRACT
This chapter provides a brief overview of ordinary regression and includes the SAS program for building and scoring an ordinary regression model. It presents a mini case study to illustrate that the data mining techniques carry over with minor modification to ordinary regression. Model builders, who are called on to provide statistical support to managers monitoring expected revenue from marketing campaigns, will find an excellent reference for profit modeling. The objective of the mini case study is as follows: to build an ordinary least squares (OLS) profit model based on INCOME and AGE. The ordinary regression model is the quintessential linear model, which implies the all-important assumption: The underlying relationship between a given predictor variable and the dependent variable is linear. In data mining, the assessment of the importance of a subset of variables for predicting profit involves the notion of a noticeable reduction in prediction error due to the subset of variables.
