Demand Forecasting with Artificial Neural Network in a Milk and Dairy Production Company

Many companies operating in the food industry relies on accurate demand forecasts to manage inventory, production planning, and ensure customer satisfaction. However, inaccurate or incomplete demand forecasting can lead to several problems in supply chain management. In competitive markets, planning and predicting the production and distribution process of these products is a factor that increases market power. Failure to predict demand correctly may lead to both excess stock accumulation and product spoilage. For this reason, it can be seen in the literature that perishable and cold chain products are being researched more and more among food products. In this study, the importance of demand forecasting in terms of the supply chain of products requiring cold chain in the food industry is discussed, and demand forecasting is improved by analyzing the past order data for 2023 of a food company with various products such as milk and dairy products. What impact this improvement will have on critical areas such as stock management, production planning, customer satisfaction, and competitiveness will be examined, and suggestions will be presented. In the application part, demand forecast results were obtained using linear regression and artificial neural network models and the two models were compared. As a result of the analysis, it was determined that the most effective improvement model was artificial neural networks and the company's forecast values were significantly improved. According to the analysis of the error values made after the application, it was observed that the predictions made by the model were reliable and consistent. This study aims to contribute to research on demand forecasting in the food industry. Additionally, showing that artificial intelligence algorithms and approaches are practically applicable in this field is one of the most important contributions of this study.