Title: Supply chain inventory stockout prediction using machine learning classifiers
Authors: Dony S. Kurian; C.R. Maneesh; V. Madhusudanan Pillai
Addresses: Department of Mechanical Engineering, National Institute of Technology Calicut, NIT Campus P.O., Calicut – 673601, Kerala, India ' Department of Mechanical Engineering, National Institute of Technology Calicut, NIT Campus P.O., Calicut – 673601, Kerala, India ' Department of Mechanical Engineering, National Institute of Technology Calicut, NIT Campus P.O., Calicut – 673601, Kerala, India
Abstract: Inventory stockouts occurring in supply chain systems are expensive and it is common in distribution systems. Nowadays, organisations are interested in making use of predictive inventory analytics to reduce stockouts and thereby achieving competitive advantage. Predicting the periods where stockout occurs will help the organisations to take preventive measures and to improve the overall supply chain performance. In this paper, machine learning classifiers are used to predict the occurrence of stockout in a period, and are proposed for the members of a four-stage serial supply chain that operates on order-up-to (OUT) inventory policy. Initially, the classifier models are trained and tested using the data collected through a spreadsheet-based simulation experiment, and performance of the classifiers are then evaluated. Application of the prediction model shows that the supply chain operated using OUT policy and stockout prediction outperforms the same supply chain operated using OUT policy alone.
Keywords: supply chain; inventory stockout prediction; classification; machine learning; ensemble methods.
DOI: 10.1504/IJBDA.2020.108700
International Journal of Business and Data Analytics, 2020 Vol.1 No.3, pp.218 - 231
Received: 30 Apr 2019
Accepted: 28 Sep 2019
Published online: 27 Jul 2020 *