The paper presents an approach to construction of hierarchical structures of data based concepts (granules), extending the idea of feedforward neural networks. The operations of processing the concept information and changing the concept specification through the network layers are discussed. Examples of the concepts and their connections are provided with respect to the case study of learning hierarchical rule based classifiers from data. The proposed methods are referred to the foundations of granular and rough-neural computing.
CITATION STYLE
Ślȩzak, D., Szczuka, M., & Wróblewski, J. (2005). Feedforward concept networks. In Advances in Soft Computing (Vol. 28, pp. 280–292). https://doi.org/10.1007/3-540-32370-8_21
Mendeley helps you to discover research relevant for your work.