A heuristic neural network model in the research of properties of evolutionary trajectories

Abstract

There is considerable data on molecular evolution, but there remains no approach to systematizing them within the framework of the key problems of biology. To search for the most common properties of evolving systems, the heuristic method has been proposed. Artificial networks of formal neurons were chosen as the heuristic model object. The paper examines the divergent component of evolutionary trajectory formation. As a result of the simulation, the dependence of the potential variability parameter on the position of the fitness function landscape was obtained. The simulation results are in agreement with the real data of molecular evolution experiments.