Modeling of computational perception of reality, situational awareness, cognition and machine learning under uncertainty

Abstract

The paper suggests how to model a computational perception of reality, situational awareness, cognition and machine learning, in the system of Computational Systemic Deep Mind. The modules of this model, determine the states of objects in the environment of unknown in advance situation, represent and realize the objects by machine memory, display the objects to be recognized by systems and humans. The method applies to the principles of the systemic and situational control, the main achievements of fuzzy logic, linguistics and cyber-physical approach to the perception, understanding and processing of languages, images, signals and other essences of reality. The functionality of this method is based on the following interconnected modules: 1) situational fuzzy control of data, information, knowledge, objects and subsystems; 2) fuzzy inference; 3) decisions making; 4) knowledge representation; 5) knowledge generalization; 6) reasoning; 7) systems thinking; and 8) intelligent user interface. The use of this method allows to be self-organized under uncertainty and to operate autonomously in various subject areas.