The experiment presented in this paper has used an unsupervised learning technique to forecast online purchasing based on historic in-store data. The methodology is an innovative software tool called LAMDA (Aguilar-Martin and López de Mántaras, 1982; Aguilar-Martin and Piera, 1986; Aguado, 1998) based on the fuzzy concept of adequacy (Aguado, 1998; Casabayó et al., 2004). Assumed the fact that online purchasing is mainly motivated by shopping convenience, the paper describes how this approach is capable to help retailers to forecast the current customers who are going to buy online. From a managerial perspective, a more realistic way to interpret the results to support decision making in marketing has been introduced as it is capable to deal with ambiguous, uncertain and incomplete information.

<< Back to home