Machine and Deep Learning Applications for Purchasing Plan

The fashion retail industry is dynamic and unpredictable, making accurate demand forecasting crucial for developing successful purchasing strategies. This thesis addresses the challenges of creating an effective purchasing plan for a footwear company in the fashion sector, where demand is highly erratic and frequently fluctuates, making accurate forecasting. To manage these challenges, the study uses advanced machine learning and deep learning algorithms to develop a forecasting model suited to the unpredictable nature of fashion demand. The research compares different models to evaluate their effectiveness in guiding purchasing decisions. The findings show that deep learning models, when fine-tuned, can significantly improve forecast accuracy, leading to a more responsive and efficient purchasing plan. This thesis helps to better understand how modern forecasting techniques can improve purchasing processes in the fashion retail industry, helping companies manage the complexities of erratic demand.