Inventory decisions under stochastic demand scenario with high inflation rate-ml approach

Abstract

Hyperinflation is a situation where prices increase at an average monthly rate of 50% or more, leading to a rapid loss of the currency’s value and causing severe economic problems. Inventory decisions under hyperinflation are crucial due to the high level of uncertainty and the rapid increase in prices can lead to significant losses if inventory is not properly managed. We examine the effects of hyperinflation on inventories of Biscuits and develop an ML model to forecast optimal order quantities of Biscuit products, with the intention of lowering inventory holding costs and inventory deterioration. Data from a retail company in Sri Lanka during the hyperinflation period of 2022 have been used to develop the ML model to predict customer demand. Six ML techniques were utilized to achieve the research objectives. Root Mean Squared Error (RMSE) and R-squared metrics are employed to choose the best model. We find that Random Forest is the most appropriate ML model to forecast optimal order quantities during a hyperinflation situation. The outcomes of our study will aid professionals working in the Biscuit industry to effectively handle inventory management during periods of hyperinflation. Our ML model can serve as a fundamental tool for predicting inventory levels during hyperinflation, which can be used as a starting point for further analysis.