Transfer Learning for Business Cycle Identification

Marcelle Chauvet and Rafael Guimaraes

No 545, Working Papers Series from Central Bank of Brazil, Research Department

Abstract: A transfer learning strategy is proposed to identify business cycles phases when data are limited or there is no business cycle dating committee. The approach integrates the idea of storing knowledge gained from one region’s economics experts and applying it to other geographic areas. The first is captured with a supervised deep neural network model, and the second by applying it to another dataset, a domain adaptation procedure. The results indicate the method proposed leads to successful business cycle identification.

Date: 2021-02
New Economics Papers: this item is included in nep-big and nep-cwa
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (6)

Downloads: (external link)
https://www.bcb.gov.br/content/publicacoes/WorkingPaperSeries/wps545.pdf (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:bcb:wpaper:545

Access Statistics for this paper

More papers in Working Papers Series from Central Bank of Brazil, Research Department
Bibliographic data for series maintained by Rodrigo Barbone Gonzalez ().