ScholarLens Back to search
1987•Journal of the Royal Statistical Society Series C (Applied Statistics)Requires access

Robust Identification of Autoregressive Moving Average Models

Guido Masarotto

Open publisher page 14 citations

Abstract

SUMMARY We introduce a class of robust estimates for the partial autocorrelation function of a univariate stationary time series and show that it is possible to produce an estimate of the autocorrelation function from the estimated partial autocorrelation coefficients. These statistics seem suitable for the preliminary identification of the order, p and q of an ARMA (p, q) model when the observed series contains a few aberrant observations.

About this research paper

What this paper is about

SUMMARY We introduce a class of robust estimates for the partial autocorrelation function of a univariate stationary time series and show that it is possible to produce an estimate of the autocorrelation function from the estimated partial autocorrelation coefficients. These statistics seem suitable for the preliminary identification of the order, p and q of an ARMA (p, q) model when the observed series contains a few aberrant observations.

Why it matters

OpenAlex reports 14 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

SUMMARY We introduce a class of robust estimates for the partial autocorrelation function of a univariate stationary time series and show that it is possible to produce an estimate of the autocorrelation function from the estimated partial autocorrelation coefficients. These statistics seem suitable for the preliminary identification of the order, p and q of an ARMA (p, q) model when the observed series contains a few aberrant observations.

Key concepts: Autoregressive model, Autoregressive–moving-average model, Identification (biology), Econometrics, Autoregressive integrated moving average, Mathematics, STAR model, Statistics

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
Robust Identification of Autoregressive Moving Average Models — Research Paper | ScholarLens