Forecasting with Multivariate Threshold Autoregressive Models

Q3 Mathematics

Revista Colombiana De Estadistica Pub Date : 2021-07-12 DOI:10.15446/RCE.V44N2.91356

Sergio Calderon, Fabio H. Nieto

引用次数: 0

Abstract

An important stage in the analysis of time series is the forecasting. How- ever, the forecasting in non-linear time series models is not straightforward as in linear time series models because an exact analytical of the conditional expectation is not easy to obtain. Therefore, a strategy of forecasting with multivariate threshold autoregressive(MTAR) models is proposed via predictive distributions through Bayesian approach. This strategy gives us the forecast for the response and exogenous vectors. The coverage percentages of the forecast intervals and the variability of the predictive distributions are analysed in this work. An application to Hydrology is presented.

查看原文本刊更多论文

多元阈值自回归模型预测

时间序列分析的一个重要阶段是预测。然而，非线性时间序列模型中的预测并不像线性时间序列模型那样简单，因为不容易获得条件期望的精确分析。因此，通过贝叶斯方法的预测分布，提出了一种多元阈值自回归（MTAR）模型的预测策略。这一策略为我们提供了对反应和外源载体的预测。本文分析了预测区间的覆盖率和预测分布的可变性。介绍了在水文学中的应用。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Revista Colombiana De Estadistica STATISTICS & PROBABILITY-

CiteScore

1.20

自引率

0.00%

发文量

审稿时长

>12 weeks

期刊介绍： The Colombian Journal of Statistics publishes original articles of theoretical, methodological and educational kind in any branch of Statistics. Purely theoretical papers should include illustration of the techniques presented with real data or at least simulation experiments in order to verify the usefulness of the contents presented. Informative articles of high quality methodologies or statistical techniques applied in different fields of knowledge are also considered. Only articles in English language are considered for publication. The Editorial Committee assumes that the works submitted for evaluation have not been previously published and are not being given simultaneously for publication elsewhere, and will not be without prior consent of the Committee, unless, as a result of the assessment, decides not publish in the journal. It is further assumed that when the authors deliver a document for publication in the Colombian Journal of Statistics, they know the above conditions and agree with them.