Saved in:
Bibliographic Details
Main Authors: Bai, Zong-Han, Chu, Po-Yen
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.12749
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912994156347392
author Bai, Zong-Han
Chu, Po-Yen
author_facet Bai, Zong-Han
Chu, Po-Yen
contents Intermittent demand forecasting poses unique challenges due to sparse observations, cold-start items, and obsolescence. Classical models such as Croston, SBA, and the Teunter--Syntetos--Babai (TSB) method provide simple heuristics but lack a principled generative foundation. We introduce TSB-HB, a hierarchical Bayesian extension of TSB. Demand occurrence is modeled with a Beta--Binomial distribution, while nonzero demand sizes follow a Log-Normal distribution. Crucially, hierarchical priors enable partial pooling across items, stabilizing estimates for sparse or cold-start series while preserving heterogeneity. This framework provides a coherent generative reinterpretation of the classical TSB structure. On the UCI Online Retail dataset, TSB-HB achieves the lowest RMSE and RMSSE among all baselines, while remaining competitive in MAE. On a 5,000-series M5 sample, it improves MAE and RMSE over classical intermittent baselines. Under the calibrated probabilistic configuration, TSB-HB yields competitive pinball loss and a favorable sharpness--calibration tradeoff among the parametric baselines reported in the main text.
format Preprint
id arxiv_https___arxiv_org_abs_2511_12749
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Taxonomy-Conditioned Hierarchical Bayesian TSB Models for Heterogeneous Intermittent Demand Forecasting
Bai, Zong-Han
Chu, Po-Yen
Machine Learning
Intermittent demand forecasting poses unique challenges due to sparse observations, cold-start items, and obsolescence. Classical models such as Croston, SBA, and the Teunter--Syntetos--Babai (TSB) method provide simple heuristics but lack a principled generative foundation. We introduce TSB-HB, a hierarchical Bayesian extension of TSB. Demand occurrence is modeled with a Beta--Binomial distribution, while nonzero demand sizes follow a Log-Normal distribution. Crucially, hierarchical priors enable partial pooling across items, stabilizing estimates for sparse or cold-start series while preserving heterogeneity. This framework provides a coherent generative reinterpretation of the classical TSB structure. On the UCI Online Retail dataset, TSB-HB achieves the lowest RMSE and RMSSE among all baselines, while remaining competitive in MAE. On a 5,000-series M5 sample, it improves MAE and RMSE over classical intermittent baselines. Under the calibrated probabilistic configuration, TSB-HB yields competitive pinball loss and a favorable sharpness--calibration tradeoff among the parametric baselines reported in the main text.
title Taxonomy-Conditioned Hierarchical Bayesian TSB Models for Heterogeneous Intermittent Demand Forecasting
topic Machine Learning
url https://arxiv.org/abs/2511.12749