Analytic Programming Powered by Distributed Self-Organizing Migrating Algorithm Application

Abstract

This paper presents an idea of new algorithm combining advantages of evolutionary algorithm and simple distributed computing to perform tasks which required many re-runs of the same program. Computing time is shorted due to elementary distribution within a number of common computers via the Internet. Progressive .NET Framework technology allowing this algorithm to run effectively and examples of possible usage are also described.The algorithm deals with a problem of synthesis of the artificial neural networks using the evolutional scanning method. The basic task to be solved is to create a symbolic regression algorithm on principles of analytic programming, which will be capable of performing a convenient neural network synthesis. The main motivation here is the computerization of such synthesis and discovering so far unknown solutions.

Adaptive Strategy for Neural Network Synthesis Constant Estimation

“…The function (7) proposed by [5] as an appropriate approximation benchmark was chosen to be approximated by the ANN. …”

Adaptive Control of Neural Network Synthesis

“…Instead of Plus, AP will jump into the subspace, in this case directly to the GFS0arg. In the GFS0arg it finds the second element, which is K. And by doing so, we get (5).…”

“…The function (6) proposed by [5] as an appropriate approximation benchmark was chosen to be approximated by the ANN. (7) is depicted as a red area which could be minimized by the process of synthesis.…”