The design of FIR and IIR filters which are equivalent to a least-squares process

A procedure for designing a finite-impulse-response (FIR) filter which yields results identical to a moving least squares estimation procedure is presented. The FIR method is computationally more efficient than the least-squares process. The structure of the filter weights of the FIR filter allows one to construct an infinite-impulse-response (IIR) filter, which is even more computationally efficient.<
>