Fuzzy Computing in a MultiPurpose Neural Network Implementation

  • Authors:
  • Ciprian-Daniel Neagu;Vasile Palade

  • Affiliations:
  • -;-

  • Venue:
  • Proceedings of the 6th International Conference on Computational Intelligence, Theory and Applications: Fuzzy Days
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

A family of connectionist tools is developed as a collection of fuzzy processing operators to model logic oriented computation of fuzzy sets. A generalized neuron with fuzzy capabilities (developed using the MAPI structure proposed in [4]) could be a useful tool to add another level of programmability. These combinations of generalized fuzzy computation, the expanded MAPI model and specific distributed architecture, are used as a powerful processing tool in financial forecasting, particularly in a portfolio problem. The neural reasoning engine is accorded to fuzzy rules, which model a real world portfolio evaluation process, taken from the current Romanian financial context.