Preregistration: evaluating intermittent demand forecasts on five public datasets
Year: 2026
Type: article
Abstract: Preregistration of a study on the evaluation of intermittent demand forecasts on five public datasets (Auto, RAF, M5, UCI Online Retail and a sensitivity variant of it), deposited before any test window was read. It states the hypotheses, the methods compared, the scores, the stock simulation, the decision rules and the controls. PREREG.md is the protocol. PREREG_FROZEN.json holds the SHA-256 of the protocol, of every code file, of the settings chosen on development data and of each dataset description, together with the package versions. The code is at https://github.com/Causal64/intermittent-study, frozen at commit 78e7b469281e64af63257fb4e93562609d953c3a (tag prereg-frozen).
Author: Chadeffaud, Mahé
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7230943697
