A Data-Driven Framework for Joint Pricing and Inventory Optimization Using Machine Learning in the Newsvendor Problem

The Newsvendor Problem is a fundamental model used in inventory and pricing optimization under demand uncertainty, especially for perishable and short-lifecycle products. Traditionally, analytical solutions to this problem rely on predefined demand distributions, which can be restrictive real-world applications. This study proposes a data-driven approach that integrates machine learning techniques to enhance demand forecasting and optimize pricing and inventory decisions simultaneously. By eliminating the need for distributional assumptions, the proposed model leverages historical data, price sensitivity, seasonality, and other contextual factors to predict demand more accurately. The research also presents a framework for joint pricing and inventory optimization that incorporates feedback loops for iterative learning and decision refinement. The results indicate that integrating machine learning into the Newsvendor Problem significantly enhances the ability to adapt to dynamic market conditions, optimize expected profits, and minimize costs related to inventory shortages and overstocking. This approach provides a more robust and flexible solution compared to traditional methods.

کلیدواژه ها:

Newsvendor Problem ، Machine Learning ، Inventory Optimization ، Pricing ، Artificial Intelligence