Comparision of predictors of time series in orthogonal regression models

Rok, strany: 2008, 175 - 182

The problem of comparison of the best linear predictor defined in a finite discrete spectrum model, and of the best linear unbiased predictor defined in a simple linear regression model, is considered. Mean squared errors of these two predictors are computed and compared in both these models under the assumption that functions generating these two models are the same and that the corresponding vectors in these models for a finite observation are orthogonal.

ISO 690:

Štulajter, F. 2008. Comparision of predictors of time series in orthogonal regression models. In Tatra Mountains Mathematical Publications, vol. 39, no.1, pp. 175-182. 1210-3195.

APA:

Štulajter, F. (2008). Comparision of predictors of time series in orthogonal regression models. Tatra Mountains Mathematical Publications, 39(1), 175-182. 1210-3195.