Sébastien Thomassey

ENSAIT – Lille 1 University Roubaix – France

Abstract The fashion industry is a very challenging market for the sales forecasting. Indeed, the volatile demand, the strong seasonality of sales, the wide number of items with short life cycle or the lack of historical data, make the traditional forecasting models, based on time series, unsuitable or at least very difficult to implement. Therefore, researchers and practitioners have investigated other techniques such as Artificial Intelligence. The emergence of the “big data” era has also opened new opportunities and generated new issues. In this presentation, we propose an overview of the different constraints related to the sales forecasting in the fashion industry. Some examples are introduced to illustrate the problematics and the potential solutions. Finally, the new opportunities which could be explored are discussed.