No label defined (Q4955678)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1449646
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 1449646 |
Statements
21 May 2000
0 references
locally weighted regression
0 references
time-series decomposition
0 references
quantile autoregression prediction
0 references
0.7962095737457275
0 references
0.7883752584457397
0 references
0.768906831741333
0 references
0.7556264996528625
0 references
0.753665030002594
0 references
The paper consists of two parts. In the first part, a nonparametric smoothing technique is considered for time-series quantile estimation. The kernel estimators for quantiles, quantile autoregressive forecasts, cross-validation for bandwidth selection and goodness of fit measures for forecasting are discussed. The kernel smoother is applied to the analysis of DAX-returns (Jan 1986 -- Aug 1993) analysis. The results are compared to GARCH estimates. In the second part a locally weighted regression is used for the decomposition of time series into trend-cyclical and seasonal components. The seasonal component is modeled by a local Fourier series and the trend component by a local Taylor expansion. The coefficients of the series are estimated by locally weighted least squares. Optimal bandwidth selection for the weights is discussed. This technique is applied to the decomposition of the quarterly gross domestic product in Germany (1968 -- 1994). For the entire collection see [Zbl 0929.00046].
0 references