Supply Chain Modelling Based on Twelve Related Features: A Novel Iteration Feature Selection Method

Abstract

Real-time prediction of hour-based order entry has been lacking in literature. Compared to previous research on supply chain problems, our proposed approach overcomes the constraints of operations management with longer time periods such as weekly and monthly by developing a novel iteration model. We performed experiments on 100 products with high cumulative volume over time. Using 3 different dataset, our proposed model proved efficient in forecasting skewed demand signals with lot of noise in supply chains.