Using FCA for Encoding Closure Operators into Neural Networks

  • Authors:

  • Sebastian Rudolph

  • Affiliations:

  • University of Karlsruhe, Institute AIFB,

  • Venue:
  • ICCS '07 Proceedings of the 15th international conference on Conceptual Structures: Knowledge Architectures for Smart Applications
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

After decades of concurrent development of symbolic and connectionist methods, recent years have shown intensifying efforts of integrating those two paradigms. This paper contributes to the development of methods for transferring present symbolic knowledge into connectionist representations. Motivated by basic ideas from formal concept analysis, we propose two ways of directly encoding closure operators on finite sets in a 3-layered feed forward neural network.