Stepwise Regression

Algorithm

Stepwise regression is used when there is uncertainty about which of a set of predictor variables should be included in a regression model. It works by adding and/or removing individual variables from the model and observing the resulting effect on its accuracy.

Stepwise regression is no longer regarded as a valid tool for dimensionality reduction because it produces unstable results that heavily overfit the training data, but see least angle regression (LARS).

alias
subtype
has functional building block
FBB_Dimensionality reduction
has input data type
IDT_Vector of quantitative variables
has internal model
INM_Function
has output data type
has learning style
LST_Supervised
has parametricity
PRM_Parametric
has relevance
REL_Obsolete
uses
ALG_Least Squares Regression ALG_Logistic regression
sometimes supports
ALG_Least Squares Regression ALG_Logistic regression
mathematically similar to