InformeDesign: Glossary of Terms

Backward-elimination
(backward-elimination)

A method of selecting variables for a regression model that eliminates variables that do not make a significant contribution to prediction by selecting those variables that will work best first.