A new algorithm for ARMA model parameter estimation using group method of data handling

Author

Lu, S. ; Chon, K.H.

Author_Institution

Dept. of Electr. Eng., City Coll. of New York, NY, USA

fYear

2000

fDate

2000

Firstpage

127

Lastpage

128

Abstract

A new method for autoregressive moving average (ARMA) parameter estimation is introduced. The algorithm is based on the Group Method of Data Handling (GMDH) first introduced by A.G. Ivakhnenko (1966, 1971), for solving high-order regression polynomials. We modified the GMDH algorithm to solve for ARMA model parameters. Computer simulations show that in cases with noise contamination and incorrect model order assumptions, the GMDH usually performs better than either the FOS or the least-squares methods in providing only the parameters that are associated with the true model terms

Keywords

autoregressive moving average processes; forecasting theory; iterative methods; parameter estimation; physiological models; polynomials; ARMA model parameter estimation; computer simulations; data partitioning; group method of data handling; high-order regression polynomials; incorrect model order assumptions; least square errors; linear ARMA model; modified algorithm; noise contamination; physiological system identification; true model terms; Algorithm design and analysis; Autoregressive processes; Biomedical engineering; Cities and towns; Data handling; Equations; Least squares methods; Parameter estimation; Polynomials; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Bioengineering Conference, 2000. Proceedings of the IEEE 26th Annual Northeast

Conference_Location

Storrs, CT

Print_ISBN

0-7803-6341-8

Type

conf

DOI

10.1109/NEBC.2000.842412

Filename

842412