| dc.contributor.author | Sinha, Sanjoy Kumar. | en_US |
| dc.date.accessioned | 2014-10-21T12:35:19Z | |
| dc.date.available | 2000 | |
| dc.date.issued | 2000 | en_US |
| dc.identifier.other | AAINQ57354 | en_US |
| dc.identifier.uri | http://hdl.handle.net/10222/55703 | |
| dc.description | In this thesis, a number of robust methods have been developed for estimating the parameters in a time series setting. To estimate the power spectrum of an ARMA process, an M estimation method has been introduced which maximizes the robust likelihood function of the discrete Fourier transforms of the process. This robust method is useful in estimating the parameters of the continuous spectrum ARMA process by downweighting the influence of possible discrete spectrum harmonic components on the data. The proposed M estimation method has been applied to some actual time series data sets of sea level records, where a strong presence of tidal (harmonic) components is observed along with the continuous spectrum surge process. Here robust estimation of the power spectrum of the surge process has been considered assuming that the surge follows an ARMA process. | en_US |
| dc.description | A GM estimation technique has been introduced for the robust estimates of the parameters in a nonlinear regression setting with autoregressive errors. The asymptotic properties of the GM estimators have been studied in some detail. For choosing the appropriate order of an autoregressive process, a robust criterion has also been suggested, which uses the robust version of the Akaike Information Criterion. The proposed GM estimation and model selection criterion have been applied to a ozone data set which appears to have nonlinear relationship with some meteorological variables. As the data are collected sequentially over time, there appears to be a significant serial correlation in the errors. A nonlinear regression model with autoregressive errors has been fitted to this data set for the joint estimates of the regression parameters and the autoregressive parameters. | en_US |
| dc.description | A new class of influence functions in the frequency domain has been introduced in this thesis. Like Hampel's influence function, this frequency influence function appears to have some close relationship with the asymptotic variance of a time series functional. | en_US |
| dc.description | Thesis (Ph.D.)--Dalhousie University (Canada), 2000. | en_US |
| dc.language | eng | en_US |
| dc.publisher | Dalhousie University | en_US |
| dc.publisher | | en_US |
| dc.subject | Statistics. | en_US |
| dc.title | Some aspects of robust estimation in time series analysis. | en_US |
| dc.type | text | en_US |
| dc.contributor.degree | Ph.D. | en_US |
