Open this publication in new window or tab >>2019 (English)In: 31st Nofoma Conference, Supply Chains and Sustainable Development of Societies, Norwegian Business School, Oslo, June 12-14, 2019, 2019Conference paper, Oral presentation with published abstract (Refereed)
Abstract [en]
Purpose
In fast moving consumer goods sector (FMCGs), manufacturers’ access to demand related data (e.g. point of sales data, lost sales, end-customer orders) is often limited making demand forecasting impossible. Instead, to enhance production planning, the manufacturers might base their DC supply forecasts on limited information including shipments from DC to wholesalers and retail stores. This is not optimal. To enhance the situation, the purpose is: (i) to propose a method of how manufacturers can use Machine Learning (ML) approach to improve supply forecasting to a DC if demand information is lacking, and (ii) to identify the demand information needed for demand forecasting supported by ML approach.
Design/methodology/approach
A single case study methodology of a manufacturer from FMCG sector, principle component analysis, and simple linear regression in R programming language is used. Also, linear regression using stochastic gradient descent optimization technique in Amazon ML is applied.
Findings
We propose a method of how FMCG manufacturers lacking demand data can utilize the ML approach to enhance supply forecasting, and identify needed demand related information to implement demand forecasting.
Research limitations/implications (if applicable)
The results are based on one case and need validation through more cases.
Practical limitations/implications (if applicable)
The results can be used by managers lacking demand related data to improve DC supply forecasting in FMCGs sector. Originality/valueThe results can contribute to the area of supply forecasting through ML.
Keywords
FMCGs, forecasting, distribution center, machine learning, principal component analysis
National Category
Transport Systems and Logistics
Research subject
Economy, Logistics; Statistics/Econometrics
Identifiers
urn:nbn:se:lnu:diva-120282 (URN)
Conference
31st Nofoma Conference, Supply Chains and Sustainable Development of Societies, Norwegian Business School, Oslo, June 12-14, 2019
2023-04-162023-04-162025-05-07Bibliographically approved