Gorbunov I.V.   Sarin C.S.  

The controller multithreading optimization fuzzy approximation

Reporter: Gorbunov I.V.

    Fuzzy approximation could be implemented as parallel computation form. Each approximation fuzzy rule could be computed independently from others [1].

    Fuzzy approximates built by machine learning algorithms. Time for build each fuzzy approximate by these algorithms is considerable. Parallel versions of optimization algorithms exist or are being developed. The simultaneous execution a group of parallel version algorithms is required consider mutual influence their calculation on the performance all computation process.

    In the paper proposed structure of controller multithreaded optimization based on the original rules and way of automatic generation of these rules. The conditional part of the rules manager specified manufacturer, processor family architecture x86 (x86-64) and type of optimization.

    The project has been supported by the Ministry of Education and Science of the Russian Federation in the framework of the base part of state order TUSUR 2015 (project № 3657).

1. Nojima Y., Mihara S., Ishibuchi H. Parallel Distributed Implementation of Genetics-Based Machine Learning for Fuzzy Classifier Design / // Simulated Evolution and Learning. 2010. V. 6457. P. 309-318.

To reports list