Methods for recursive robust estimation of AR parameters

Sejling, Ken; Madsen, Henrik; Holst, Jan; Holst, Ulla; Englund, Jan‐Eric

doi:10.1016/0167-9473(94)90146-5

Cited by 10 publications

(2 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, a robust procedure for recursive estimation of an autoregressive process which does not depend on a priori knowledge of the scale of the observations was proposed by Sejling et al (1990). The method is based on simultaneous M-estimation of a regression but makes use of certain local approximations which enable the parameter estimates to be recursively computed.…”

Section: Introductionmentioning

confidence: 99%

“…This type of non-stationarity is of particular interest here, since the procedure is expected to separate the effects of outliers from system changes. However, the simulation results presented by Sejling et al (1990) are restricted to the stationary case and, although a recursive estimator is proposed, it was not used in the simulation study. Clearly these aspects are important and require investigation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Robust Methods for Recursive Autoregressive Moving Average Estimation

McDougall

1994

Journal of the Royal Statistical Society Series B: Statistical Methodology

View full text Add to dashboard Cite

In this paper we consider the problem of fitting, recursively in time, an autoregressive moving average (ARMA) model to time series data where outliers may be present. Although many recursive estimation procedures are available for fitting ARMA models, they are based on the recursive least squares algorithm which is known to be badly affected by additive outliers. To minimize the impact of these sort of outliers we investigate some robustified versions of the recursive maximum likelihood procedure. The problem of incorporating robust scale recursion is discussed and asymptotic properties of the procedures are examined. We also present results from simulation studies which compare adaptive versions of these procedures in the non-stationary case.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%