Skip navigation
PDF 146 KB
DOI: http://dx.doi.org/10.7551/978-0-262-32621-6-ch069
Pages 427-428
First published 30 July 2014

Random Fuzzy Networks

Octavio B. Zapata and Carlos Gershenson

Abstract (Excerpt)

No model can be considered effective if fundamentally it is more complicated than what it’s trying to represent. However, extreme simplification may potentially overlook important non-primary features, or even neglect the possibility to represent ambiguous or unclear observations. Therefore, achieving a balance between parsimonious and detailed models is of utmost importance for science and engineering.

Over the past few decades, random Boolean networks (RBNs) (Kauffman, 1969) have become popular models for genetic regulatory networks. This popularity is associated with the fact that RBNs are very general models. No functionality or structure is particularly assumed when constructing them. However, the Boolean idealization has been constantly criticized based on the assumption that constraining the variables of the model to have only two possible values (0 and 1) entails a loss of dynamical information in the analysis of real gene expression data.