Title: Neural networks based vendor-managed forecasting: a case study

Authors: Atul B. Borade, Satish V. Bansod

Addresses: Mechanical Engineering Department, Jawaharlal Darda Institute of Engineering and Technology, Yavatmal, Maharashtra, India. ' Mechanical Engineering Department, Professor Ram Meghe Institute of Technology and Research, Badnera, Maharashtra, India

Abstract: Vendor-managed inventory (VMI) is a collaborative supply chain management practice adopted by many organisations. For making inventory-related decisions an accurate forecast is needed. Traditional forecasting models provide close but not accurate forecasts. In the recent years, decision support tools, like neural networks, are used for making an accurate forecast. This paper presents a case study of a small enterprise where a vendor-managed inventory pact was in force between enterprise and a retailer. In the study, various neural networks were used for demand forecasting. The results of neural network based forecasts are found and compared on various fronts. Multi-criteria decision-making tools are adopted for comparing and verifying the results. Study shows that even small enterprise could adopt the simple VMI system by using properly trained neural network and obtain substantial saving in inventory and costs.

Keywords: neural networks; VMI; vendor managed inventories; accuracy; multicriteria decision making; MCDM; collaborative supply chains; collaboration; accurate forecasts; small and medium-sized enterprises; SMEs; inventory pacts; retailers; demand forecasting; comparisons; verification; cost savings; integrated supply chains; SCM; supply chain management.

DOI: 10.1504/IJISM.2011.040713

International Journal of Integrated Supply Management, 2011 Vol.6 No.2, pp.140 - 164

Received: 10 Feb 2010
Accepted: 14 Dec 2010
Published online: 17 Jun 2011 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article