Enregistré dans:
Détails bibliographiques
Auteurs principaux: Sprangers, Olivier, Wadman, Wander, Schelter, Sebastian, de Rijke, Maarten
Format: Preprint
Publié: 2023
Sujets:
Accès en ligne:https://arxiv.org/abs/2310.12809
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866929255512801280
author Sprangers, Olivier
Wadman, Wander
Schelter, Sebastian
de Rijke, Maarten
author_facet Sprangers, Olivier
Wadman, Wander
Schelter, Sebastian
de Rijke, Maarten
contents Existing hierarchical forecasting techniques scale poorly when the number of time series increases. We propose to learn a coherent forecast for millions of time series with a single bottom-level forecast model by using a sparse loss function that directly optimizes the hierarchical product and/or temporal structure. The benefit of our sparse hierarchical loss function is that it provides practitioners a method of producing bottom-level forecasts that are coherent to any chosen cross-sectional or temporal hierarchy. In addition, removing the need for a post-processing step as required in traditional hierarchical forecasting techniques reduces the computational cost of the prediction phase in the forecasting pipeline. On the public M5 dataset, our sparse hierarchical loss function performs up to 10% (RMSE) better compared to the baseline loss function. We implement our sparse hierarchical loss function within an existing forecasting model at bol, a large European e-commerce platform, resulting in an improved forecasting performance of 2% at the product level. Finally, we found an increase in forecasting performance of about 5-10% when evaluating the forecasting performance across the cross-sectional hierarchies that we defined. These results demonstrate the usefulness of our sparse hierarchical loss applied to a production forecasting system at a major e-commerce platform.
format Preprint
id arxiv_https___arxiv_org_abs_2310_12809
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Hierarchical Forecasting at Scale
Sprangers, Olivier
Wadman, Wander
Schelter, Sebastian
de Rijke, Maarten
Machine Learning
Existing hierarchical forecasting techniques scale poorly when the number of time series increases. We propose to learn a coherent forecast for millions of time series with a single bottom-level forecast model by using a sparse loss function that directly optimizes the hierarchical product and/or temporal structure. The benefit of our sparse hierarchical loss function is that it provides practitioners a method of producing bottom-level forecasts that are coherent to any chosen cross-sectional or temporal hierarchy. In addition, removing the need for a post-processing step as required in traditional hierarchical forecasting techniques reduces the computational cost of the prediction phase in the forecasting pipeline. On the public M5 dataset, our sparse hierarchical loss function performs up to 10% (RMSE) better compared to the baseline loss function. We implement our sparse hierarchical loss function within an existing forecasting model at bol, a large European e-commerce platform, resulting in an improved forecasting performance of 2% at the product level. Finally, we found an increase in forecasting performance of about 5-10% when evaluating the forecasting performance across the cross-sectional hierarchies that we defined. These results demonstrate the usefulness of our sparse hierarchical loss applied to a production forecasting system at a major e-commerce platform.
title Hierarchical Forecasting at Scale
topic Machine Learning
url https://arxiv.org/abs/2310.12809