Using Neural Networks for Clustering-Based Market Segmentation

Abstract: We study use of artificial neural networks in clustering-based market segmentation. To this end two types of feedforward neural networks with logistic activation functions are formulated. Models of the first type determine segments on the basis of segmentation criteria. Models of the second type simultaneously form segments and discriminate between these segments on the basis of additional segment descriptors. Parameters of all models are estimated by an extended version of backpropagation.;