The Future of Forecasting Competitions: Design Attributes and Principles
@article{Makridakis2021TheFO, title={The Future of Forecasting Competitions: Design Attributes and Principles}, author={Spyros Makridakis and Chris Fry and Fotios Petropoulos and Evangelos Spiliotis}, journal={INFORMS Journal on Data Science}, year={2021} }
Forecasting competitions are the equivalent of laboratory experimentation widely used in physical and life sciences. They provide useful, objective information to improve the theory and practice of forecasting, advancing the field, expanding its usage, and enhancing its value to decision and policymakers. We describe 10 design attributes to be considered when organizing forecasting competitions, taking into account trade-offs between optimal choices and practical concerns, such as costs, as…
6 Citations
A review of machine learning concepts and methods for addressing challenges in probabilistic hydrological post-processing and forecasting
- Computer ScienceFrontiers in Water
- 2022
In this review, key ideas and information that can lead to effective popularizations are emphasized, as such an emphasis can support successful future implementations and further scientific developments.
Responses to the discussions and commentaries of the M5 Special Issue
- GeologyInternational Journal of Forecasting
- 2022
Exploring the representativeness of the M5 competition data
- BusinessInternational Journal of Forecasting
- 2021
References
SHOWING 1-10 OF 46 REFERENCES
M5 accuracy competition: Results, findings, and conclusions
- BusinessInternational Journal of Forecasting
- 2022
Cross-temporal aggregation: Improving the forecast accuracy of hierarchical electricity consumption
- Economics, Computer ScienceApplied Energy
- 2020
Investigating the accuracy of cross-learning time series forecasting methods
- Computer ScienceInternational Journal of Forecasting
- 2021