Data-driven mechanisms for a newsvendor problem: a case study

Official URL: https://dx.doi.org/10.35378/gujs.1334184

Reducing food waste is paramount for a sustainable future as its implications are important to achieving sustainable development goals set by the United Nations. In many industry groups, the public awareness of reducing food waste that may potentially emerge along firms’ operations has grown. In the era of Big Data, one of the most pursued exercises of this escalating attention on reducing food waste is to utilize artificial intelligence techniques to incorporate sustainability concerns into the decision framework. Many firms embrace machine learning methods to build effective decision mechanisms that help make efficient and sustainable decisions. In this study, we analyze the impact of blending machine learning approaches with demand forecasting and order quantity decisions for a firm operating in a setting where the market demand is random, and the demand structure is not observable to the firm. The performance of the methodology is evaluated on sunflower seed demand data taken from Tadım company. Our results suggest that the joint consideration of forecasting and ordering decisions using the quantile regression approach can lead the firm to decrease its operational cost by 8,11% on average.