Stepwise Regression

Rather than start with every available explanatory variable (which forces us to exclude variables like Plant that are redundant), forward stepwise regression builds a regression from the ground up, adding one variable to the fit at a time. The algorithm is straightforward.

  1. Initialization. Stepwise regression begins with an initial model. This model usually has no explanatory variables, but you can force the initial model to include certain explanatory variables. For example, we might force the regression to use the explanatory variable 1/Units to require an estimate of fixed costs.

  2. Search. The algorithm searches the collection of potential ...

Get Statistics for Business: Decision Making and Analysis, 3rd Edition now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.