Home > Research > Publications & Outputs > Cross-validation aggregation for combining auto...

Electronic data

  • Postprint

    Rights statement: This is the author’s version of a work that was accepted for publication in International Journal of Forecasting. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in International Journal of Forecasting, 32, 4, 2016 DOI: 10.1016/j.ijforecast.2015.12.011

    Accepted author manuscript, 1 MB, PDF-document

    Available under license: CC BY-NC-ND: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License

Links

Text available via DOI:

View graph of relations

Cross-validation aggregation for combining autoregressive neural network forecasts

Research output: Contribution to journalJournal article

Published
<mark>Journal publication date</mark>1/10/2016
<mark>Journal</mark>International Journal of Forecasting
Issue number4
Volume32
Number of pages18
Pages (from-to)1120-1137
Publication statusPublished
Early online date1/06/16
Original languageEnglish

Abstract

This paper evaluates kk-fold and Monte Carlo cross-validation and aggregation (crogging) for combining neural network autoregressive forecasts. We introduce Monte Carlo crogging which combines bootstrapping and cross-validation (CV) in a single approach through repeated random splitting of the original time series into mutually exclusive datasets for training. As the training/validation split is independent of the number of folds, the algorithm offers more flexibility in the size, and number of training samples compared to kk-fold cross-validation. The study also provides for crogging and bagging: (1) the first systematic evaluation across time series length and combination size, (2) a bias and variance decomposition of the forecast errors to understand improvement gains, and (3) a comparison to established benchmarks of model averaging and selection. Crogging can easily be extended to other autoregressive models. Results on real and simulated series demonstrate significant improvements in forecasting accuracy especially for short time series and long forecast horizons.

Bibliographic note

This is the author’s version of a work that was accepted for publication in International Journal of Forecasting. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in International Journal of Forecasting, 32, 4, 2016 DOI: 10.1016/j.ijforecast.2015.12.011