Machine learning based fuzzy inventory model for imperfect deteriorating products with demand forecast and partial backlogging under green investment technology

Abstract

This article proposes a novel approach to managing inventory by incorporating machine learning techniques to handle imperfect deteriorating products under green investment technology. The shortages are permitted and partially backlogged. Due to uncertainty, deterioration rate and defective percentage in quantity in the received lot are considered fuzzy variables. This study aims to determine optimal ordering quantity and replenishment period to minimise the total average cost with carbon emission cost. Decision Tree Classifier algorithm is used to demand forecast seasonally. The total fuzzy cost functions defuzzify by applying sign distance approach method. A numerical example is taken to illustrate the proposed model. A comparative analysis has been studied between fixed demand and month-wise forecasted demand. The study highlights the importance of forecasted demand in the inventory system and establishes methodology to get direct month-wise forecasted demand. Finally, the sensitivity analysis performs to determine more sensitive parameters and provides managerial insights.