Actor model of building neuro-fuzzy systems

Development of actors-based model of neuro-fuzzy systems is considered. Different ways of distributed asynchronous artificial intelligent systems. Actor model, which is based on an interaction of separate and independent units, allows creating of hybrid intelligent calculating systems. Properties of actors and the specifics of their communication give a number of ways for providing system’s self-diagnostics and self-organization. The comparison of efficiency of distributed neuro-fuzzy system, based on actors model, and a monolith system, created as one unit, is provided. Time of processing a specific number of datasets was performed on both if these types of systems for this purpose. Memory distribution and calculating units’ interaction of distributed hybrid artificial intelligence systems were estimated.

Authors: S. M. Morozov, M. S. Kupriyanov

Direction: Informatics, Computer Technologies And Control

Keywords: actor systems, neuro-fuzzy systems, hybrid systems, artificial intelligence, asynchronous computing

---

View full article