| Popis: |
A universal xor-metric of the digital cyberspace convolution is forming, which makes it possible to measure any cyber social processes and phenomena, as well as to determine the degree of their difference for the recognition and elimination of social collisions L based on the equation F ⊕ T ⊕ L = 0. The model of Digital Twin Computing is considered; it is characterized by comparison of two ML-truth tables that simulate the ideal and real behavior of a person with the metric grading for social activity and significance, that accepted in society. It is shown that cyber social computing makes it possible to avoid in each country multi-billion dollar losses in the economy, history, culture, science, education, by eliminating incompetent control actions that contradict centuries-old folk traditions. An advanced cloud-edge cyber social computing architecture for federated learning algorithms is proposed, which includes four phases: local training (Training), uploading parameters to the cloud model (Upload), aggregating parameters on the cloud (Aggregating), and returning model parameters to terminals (Download). A logical metric of the recognition quality of patterns, defects and collisions is introduced, which, together with the computing equation, makes it possible to form all structural and normalized assessments in the learning process. |
