“School of Cognitive%20Sciences”

Back to Papers Home
Back to Papers of School of Cognitive%20Sciences

Paper   IPM / Cognitive%20Sciences / 7537
School of Cognitive Sciences
  Title:   Multi Step-prediction of Solar Activity with ETDLMLP Ëmotional Temporal Difference Learning via Multi Layer Perception
  Author(s):  C. Lucas
  Status:   In Proceedings
  Proceeding: CIRA
  Year:  2003
  Supported by:  IPM
  Abstract:
This article presents an alternative for traditional inference mechanisms and most typically, the compositional rule of inference (CRI) method. The most important benefit of the new method is its higher robustness with respect to changes in the rule base and its ability to function when the lather is sparse. The proposed method is based on a fuzzy generalization of the ordinary least-square interpolation method. The authors consider the degradation caused by elimination of a rule in a fuzzy controller and show that CRL is more sensitive than their interpolation-based method.


Download TeX format
back to top
scroll left or right