Prof. Georgia Perakis

Online shopping has given rise to a wealth of data previously unavailable to retailers. This data has created new opportunities for companies to personalize their services to individual customers, such as through targeted promotions and personalized assortments and pricing. As a side benefit, knowledge of individual customer behavior can also help improve sales forecasting. But in order to develop consumer-targeted strategies, we first need a demand forecasting model that captures “trends” between customers (or groups of customers). Using customers’ purchase information, we have developed a machine learning algorithm that incorporates potential trends between groups of customers based on their transaction history. Unlike previous models, this model can even estimate customer demand with transaction data alone. This personalized forecasting also allows us to optimize targeted promotions to improve profits.