In-sample forecasting: a brief review and new algorithms
Statistical methods often distinguish between in-sample and out-of-sample approaches. In particular this is the case when time is involved. Then often time series methods are proposed that extrapolate past patterns into the future via complicated recursion formulas. Standard statistical inference is...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article (Journal) |
| Language: | English |
| Published: |
2018
|
| In: |
Alea
Year: 2018, Volume: 15, Issue: 2, Pages: 875-895 |
| ISSN: | 1980-0436 |
| DOI: | 10.30757/ALEA.v15-33 |
| Online Access: | Resolving-System, Volltext: http://dx.doi.org/10.30757/ALEA.v15-33 Verlag, Volltext: http://alea.impa.br/english/index_v15.htm Verlag, Volltext: http://alea.impa.br/articles/v15/15-33.pdf 
|
| Author Notes: | Y.K. Lee, E. Mammen, J.P. Nielsen and B.U. Park |
| Summary: | Statistical methods often distinguish between in-sample and out-of-sample approaches. In particular this is the case when time is involved. Then often time series methods are proposed that extrapolate past patterns into the future via complicated recursion formulas. Standard statistical inference is on the other hand concerned with estimating parameters within the given sample. This review paper is about a statistical methodology, where all parameters are estimated in-sample while producing a forecast out-of-sample without recursion or extrapolation. A new super-simulation algorithm ensures a faster implementation of the simplest and perhaps most important version of in-sample forecasting. |
|---|---|
| Item Description: | Gesehen am 17.12.2018 Der Artikel hat keine eigene Internetseite |
| Physical Description: | Online Resource |
| ISSN: | 1980-0436 |
| DOI: | 10.30757/ALEA.v15-33 |