Complex-Valued Logic for Neural Networks
Eugene Kagan, Alexander Rybalov, Ronald R. Yager · 2018
In the report, a simple complex-valued logic is derived from the algebraic structure with uninorm and absorbing norm aggregators. Such construction forms a unified framework for probabilistic and complex fuzzy logics and provides a basis for modeling decision-making in complex systems. In particular, the applicability of the obtained logic is illustrated by construction of the modules of neural networks with recursive learning.