Fuzzy model to target selection in direct marketing for large client and preferences database

Recent advances on Computational Intelligence (CI) methods have been published employing novel tools to solve problems in the world of finances, management and marketing research. Different CI techniques have been applied in data mining to obtain a model for commercial preferences of consumers and tendencies of consumption. This paper discusses and proposes an approach to model (by applying some fuzzy methods) a target selection from large databases for direct marketing. We state a methodology, via an algorithm, which pre-processes a database, by using properties of the Hotelling transformation. The processed database is used for the identification and modelling of the representative clients and preferences with a fuzzy clustering approach.