Local regression

Local regression or local polynomial regression,^[1] also known as moving regression,^[2] is a general statement of the moving average and polynomial regression.^[3]

An example of a LOESS cruve

It is used for scatterplot smoothing. They are also known as LOESS (locally estimated scatterplot smoothing) and LOWESS (locally weighted scatterplot smoothing), both pronounced /ˈloʊɛs/ LOH-ess in this care. They are two strongly related non-parametric regression methods.

In some fields, LOESS is known and commonly known as Savitzky–Golay filter.^[4][5]