Forecasting the intermittent demand for slow-moving inventories: A modelling approach

Ralph David Snyder, J Keith Ord, Adrian Nicholas Beaumont

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Organizations with large-scale inventory systems typically have a large proportion of items for which demand is intermittent and low volume. We examine various different approaches to demand forecasting for such products, paying particular attention to the need for inventory planning over a multi-period lead-time when the underlying process may be non-stationary. This emphasis leads to the consideration of prediction distributions for processes with time-dependent parameters. A wide range of possible distributions could be considered, but we focus upon the Poisson (as a widely used benchmark), the negative binomial (as a popular extension of the Poisson), and a hurdle shifted Poisson (which retains Croston s notion of a Bernoulli process for the occurrence of active demand periods). We also develop performance measures which are related to the entire prediction distribution, rather than focusing exclusively upon point predictions. The three models are compared using data on the monthly demand for 1046 automobile parts, provided by a US automobile manufacturer. We conclude that inventory planning should be based upon dynamic models using distributions that are more flexible than the traditional Poisson scheme.
Original languageEnglish
Pages (from-to)485 - 496
Number of pages12
JournalInternational Journal of Forecasting
Volume28
Issue number2
DOIs
Publication statusPublished - 2012

Cite this

@article{3a665a59d3394cc48c8797d5874708c5,
title = "Forecasting the intermittent demand for slow-moving inventories: A modelling approach",
abstract = "Organizations with large-scale inventory systems typically have a large proportion of items for which demand is intermittent and low volume. We examine various different approaches to demand forecasting for such products, paying particular attention to the need for inventory planning over a multi-period lead-time when the underlying process may be non-stationary. This emphasis leads to the consideration of prediction distributions for processes with time-dependent parameters. A wide range of possible distributions could be considered, but we focus upon the Poisson (as a widely used benchmark), the negative binomial (as a popular extension of the Poisson), and a hurdle shifted Poisson (which retains Croston s notion of a Bernoulli process for the occurrence of active demand periods). We also develop performance measures which are related to the entire prediction distribution, rather than focusing exclusively upon point predictions. The three models are compared using data on the monthly demand for 1046 automobile parts, provided by a US automobile manufacturer. We conclude that inventory planning should be based upon dynamic models using distributions that are more flexible than the traditional Poisson scheme.",
author = "Snyder, {Ralph David} and Ord, {J Keith} and Beaumont, {Adrian Nicholas}",
year = "2012",
doi = "10.1016/j.ijforecast.2011.03.009",
language = "English",
volume = "28",
pages = "485 -- 496",
journal = "International Journal of Forecasting",
issn = "0169-2070",
publisher = "Elsevier",
number = "2",

}

Forecasting the intermittent demand for slow-moving inventories: A modelling approach. / Snyder, Ralph David; Ord, J Keith; Beaumont, Adrian Nicholas.

In: International Journal of Forecasting, Vol. 28, No. 2, 2012, p. 485 - 496.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Forecasting the intermittent demand for slow-moving inventories: A modelling approach

AU - Snyder, Ralph David

AU - Ord, J Keith

AU - Beaumont, Adrian Nicholas

PY - 2012

Y1 - 2012

N2 - Organizations with large-scale inventory systems typically have a large proportion of items for which demand is intermittent and low volume. We examine various different approaches to demand forecasting for such products, paying particular attention to the need for inventory planning over a multi-period lead-time when the underlying process may be non-stationary. This emphasis leads to the consideration of prediction distributions for processes with time-dependent parameters. A wide range of possible distributions could be considered, but we focus upon the Poisson (as a widely used benchmark), the negative binomial (as a popular extension of the Poisson), and a hurdle shifted Poisson (which retains Croston s notion of a Bernoulli process for the occurrence of active demand periods). We also develop performance measures which are related to the entire prediction distribution, rather than focusing exclusively upon point predictions. The three models are compared using data on the monthly demand for 1046 automobile parts, provided by a US automobile manufacturer. We conclude that inventory planning should be based upon dynamic models using distributions that are more flexible than the traditional Poisson scheme.

AB - Organizations with large-scale inventory systems typically have a large proportion of items for which demand is intermittent and low volume. We examine various different approaches to demand forecasting for such products, paying particular attention to the need for inventory planning over a multi-period lead-time when the underlying process may be non-stationary. This emphasis leads to the consideration of prediction distributions for processes with time-dependent parameters. A wide range of possible distributions could be considered, but we focus upon the Poisson (as a widely used benchmark), the negative binomial (as a popular extension of the Poisson), and a hurdle shifted Poisson (which retains Croston s notion of a Bernoulli process for the occurrence of active demand periods). We also develop performance measures which are related to the entire prediction distribution, rather than focusing exclusively upon point predictions. The three models are compared using data on the monthly demand for 1046 automobile parts, provided by a US automobile manufacturer. We conclude that inventory planning should be based upon dynamic models using distributions that are more flexible than the traditional Poisson scheme.

U2 - 10.1016/j.ijforecast.2011.03.009

DO - 10.1016/j.ijforecast.2011.03.009

M3 - Article

VL - 28

SP - 485

EP - 496

JO - International Journal of Forecasting

JF - International Journal of Forecasting

SN - 0169-2070

IS - 2

ER -