Stacking Regression for Time-Series, with an Application to Forecasting Quarterly US GDP Growth

Erkal Ersoy, Haoyang Li, Mark E. Schaffer*, Tibor Szendrei

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Machine learning methods are being increasingly adopted in economic forecasting. Many learners are available, and a practical issue now presents itself: which one(s) to use? The answer we suggest is ‘stacking regression’ (Wolpert, 1992), an ensemble method for combining predictions of different learners. We show how to use stacking regression in the time series setting. Macroeconomic and financial time series data present their own challenges to forecasting (extreme values, regime changes, etc.), and this presents challenges to stacking as well. Our findings suggest that using absolute deviations for scoring the base learners performs well compared to stacking on mean squared error. We illustrate this with a Monte Carlo exercise and an empirical application: forecasting US GDP growth around the Covid-19 pandemic.

Original languageEnglish
Title of host publicationStudies in Systems, Decision and Control
PublisherSpringer
Pages131-149
Number of pages19
ISBN (Electronic)9783031357633
ISBN (Print)9783031357626
DOIs
Publication statusPublished - 2024

Publication series

NameStudies in Systems, Decision and Control
Volume483
ISSN (Print)2198-4182
ISSN (Electronic)2198-4190

Keywords

  • forecasting
  • machine learning
  • robust statistics
  • Stacking regression

ASJC Scopus subject areas

  • Computer Science (miscellaneous)
  • Control and Systems Engineering
  • Automotive Engineering
  • Social Sciences (miscellaneous)
  • Economics, Econometrics and Finance (miscellaneous)
  • Control and Optimization
  • Decision Sciences (miscellaneous)

Cite this