Local Regression: The LOESS Procedure

The LOESS procedure implements a local regression approach for estimating regression surfaces pioneered by Cleveland, Devlin, and Grosse (1988). No assumptions about the parametric form of the entire regression surface are made with the LOESS procedure. Only a parametric form of the local approximation is specified by the user. Furthermore, the LOESS procedure is suitable when there are outliers in the data and a robust fitting method is necessary.