Inventory Optimization for Deteriorating Products Using Exponential Demand Forecasting and Machine Learning with Carbon Emission Constraints

In today's business environment, companies implement highly advanced machine learning algorithms to calculate product demand and evaluate inventory levels. In the present research, we constructed a Degradation-based inventory model that incorporates Carbon emission, partial waiting lists, and exponentially forecasted demand. Since degradation and the rate of imperfections are both unpredictable and often difficult to accurately measure, we are treating them as fuzzy variables to represent a more realistic real-world scenario. This study proposed a method to determine the optimal order quantity and replenishment schedule that minimizes the average total cost while accounting for carbon emission expenditures. The sign distance approach is used for the defuzzification process, and demand is assumed to increase exponentially. The proposed method has been validated with numerical examples that demonstrate how demand for deteriorating products can be calculated monthly, enabling firms to optimize inventory management based on expected demand. Furthermore, a comparison between monthly predicted demand and fixed demand has been conducted, highlighting the advantages of the exponential demand model.