Using boosting for forecasting electric energy consumption during a recession: a case study for the Brazilian State Rio Grande do Sul

Guilherme Lindenmeyer, Pedro Pablo Skorin and Hudson da Silva Torrent
Additional contact information
Pedro Pablo Skorin: Universidade Federal do Rio Grande do Sul
Hudson da Silva Torrent: Universidade Federal do Rio Grande do Sul

Letters in Spatial and Resource Sciences, 2021, vol. 14, issue 2, No 2, 128 pages

Abstract: Abstract This paper seeks to test the component-wise boosting validity as an instrument of forecasting regional series in economic recessions. We use 822 predictors to forecast the monthly electricity consumption of the Brazilian state Rio Grande do Sul. The time series has 190 observations and occurs during the 2015 political and economic crisis of Brazil, the biggest economic crisis in the country’s history until then, which significantly impacted the electricity consumption behavior. Using 12 lags of the predictors, boosting manages to select predictors associated with the crisis and is capable of understanding better the trend change compared to a standard seasonal autoregressive integrated moving average (SARIMA) benchmark. We found three selected predictor clusters: lags of Y, which model the seasonality of the series, the lagged electricity consumption around Brazil, which serve to update the model the recent changes in the electricity consumption, and the unemployment rates, which are directly related to the crisis. We conclude the paper bringing two alternative exercises, using k-fold in the place of AICc and also executing a quantile-boosting exercise. By filtering a high-dimensional data set, the machine learning algorithm appears as a useful instrument when forecasting short regional series within recessions.

Keywords: Boosting; Electric energy consumption; Forecast; Regional; Recession (search for similar items in EconPapers)
JEL-codes: C53 Q47 R11 (search for similar items in EconPapers)
Date: 2021
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s12076-021-00268-3 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

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:spr:lsprsc:v:14:y:2021:i:2:d:10.1007_s12076-021-00268-3

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/12076

DOI: 10.1007/s12076-021-00268-3

Access Statistics for this article

Letters in Spatial and Resource Sciences is currently edited by Henk Folmer and Amitrajeet A. Batabyal

More articles in Letters in Spatial and Resource Sciences from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().