Suitable demand forecasting method for stock quantity optimization in the food industry

Abstract

The goal of this master's thesis is to answer the research question, which is “what is the suitable way for the food industry to make demand forecasting?”. In this research, various demand forecasting methods were compared using a dataset from the food industry. These methods included Facebook Prophet, Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN), and 90 days average, which is the research baseline. Forecasting methods are compared using Root Mean square error (RMSE) calculation. The results showed that the 90-day average method performed the best in accuracy. However, it is essential to note that other factors, such as data collection and unusual demand changes, can impact the effectiveness of demand forecasting in the food industry. The study also discussed the importance of inventory management in the food industry and the impact of stock quantity optimization, including using demand forecasts to optimize stock quantities. Overall, this research provides insights into demand forecasting in the food industry.