## FANDOM

186 Pages

Minimum mean square error (MMSE) is a signal processing criterion which attempts to minimize the mean square error of a processing output. The mean square error can be given by MSE = $E[(X - X_m)^2]$ where $X_m$ is the average of $X$ and $E[\cdot]$ is the average operator.

In particular, using the MMSE criterion the processing $w(\cdot)$ can be optimize as

$w_o = \min_{w} E{(w(X) - X_m)^2}$

where $w_o$ is the optimal processing in the sense of MMSE.