A new accuracy measure based on bounded relative error for time series forecasting

Many accuracy measures have been proposed in the past for time series forecasting comparisons. However, many of these measures suffer from one or more issues such as poor resistance to outliers and scale dependence. In this paper, while summarising commonly used accuracy measures, a special review i...

Full description

Bibliographic Details
Main Authors: Chen, Chao, Twycross, Jamie, Garibaldi, Jonathan M.
Format: Article
Published: Public Library of Science 2017
Subjects:
Online Access:https://eprints.nottingham.ac.uk/41565/
_version_ 1848796304261513216
author Chen, Chao
Twycross, Jamie
Garibaldi, Jonathan M.
author_facet Chen, Chao
Twycross, Jamie
Garibaldi, Jonathan M.
author_sort Chen, Chao
building Nottingham Research Data Repository
collection Online Access
description Many accuracy measures have been proposed in the past for time series forecasting comparisons. However, many of these measures suffer from one or more issues such as poor resistance to outliers and scale dependence. In this paper, while summarising commonly used accuracy measures, a special review is made on the symmetric mean absolute percentage error. Moreover, a new accuracy measure called the Unscaled Mean Bounded Relative Absolute Error (UMBRAE), which combines the best features of various alternative measures, is proposed to address the common issues of existing measures. A comparative evaluation on the proposed and related measures has been made with both synthetic and real-world data. The results indicate that the proposed measure, with user selectable benchmark, performs as well as or better than other measures on selected criteria. Though it has been commonly accepted that there is no single best accuracy measure, we suggest that UMBRAE could be a good choice to evaluate forecasting methods, especially for cases where measures based on geometric mean of relative errors, such as the geometric mean relative absolute error, are preferred.
first_indexed 2025-11-14T19:45:51Z
format Article
id nottingham-41565
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T19:45:51Z
publishDate 2017
publisher Public Library of Science
recordtype eprints
repository_type Digital Repository
spelling nottingham-415652020-05-04T18:38:58Z https://eprints.nottingham.ac.uk/41565/ A new accuracy measure based on bounded relative error for time series forecasting Chen, Chao Twycross, Jamie Garibaldi, Jonathan M. Many accuracy measures have been proposed in the past for time series forecasting comparisons. However, many of these measures suffer from one or more issues such as poor resistance to outliers and scale dependence. In this paper, while summarising commonly used accuracy measures, a special review is made on the symmetric mean absolute percentage error. Moreover, a new accuracy measure called the Unscaled Mean Bounded Relative Absolute Error (UMBRAE), which combines the best features of various alternative measures, is proposed to address the common issues of existing measures. A comparative evaluation on the proposed and related measures has been made with both synthetic and real-world data. The results indicate that the proposed measure, with user selectable benchmark, performs as well as or better than other measures on selected criteria. Though it has been commonly accepted that there is no single best accuracy measure, we suggest that UMBRAE could be a good choice to evaluate forecasting methods, especially for cases where measures based on geometric mean of relative errors, such as the geometric mean relative absolute error, are preferred. Public Library of Science 2017-03-24 Article PeerReviewed Chen, Chao, Twycross, Jamie and Garibaldi, Jonathan M. (2017) A new accuracy measure based on bounded relative error for time series forecasting. PLoS ONE, 12 (3). e0174202/1-e0174202/23. ISSN 1932-6203 Forecasting performance; Accuracy measure; Relative measure; Bounded error; Unscaled mean bounded relative absolute error http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0174202 doi:10.1371/journal.pone.0174202 doi:10.1371/journal.pone.0174202
spellingShingle Forecasting performance; Accuracy measure; Relative measure; Bounded error; Unscaled mean bounded relative absolute error
Chen, Chao
Twycross, Jamie
Garibaldi, Jonathan M.
A new accuracy measure based on bounded relative error for time series forecasting
title A new accuracy measure based on bounded relative error for time series forecasting
title_full A new accuracy measure based on bounded relative error for time series forecasting
title_fullStr A new accuracy measure based on bounded relative error for time series forecasting
title_full_unstemmed A new accuracy measure based on bounded relative error for time series forecasting
title_short A new accuracy measure based on bounded relative error for time series forecasting
title_sort new accuracy measure based on bounded relative error for time series forecasting
topic Forecasting performance; Accuracy measure; Relative measure; Bounded error; Unscaled mean bounded relative absolute error
url https://eprints.nottingham.ac.uk/41565/
https://eprints.nottingham.ac.uk/41565/
https://eprints.nottingham.ac.uk/41565/