SELF-ORGANIZATION PHENOMENA IN DATABASES OF QUANTITATIVE PARAMETERS OF COMPLEX SYSTEMS. FIBONACCI-LIKE STRATEGY OF REGULARITIES RECOGNITION

摘要

At certain number of observations despite the joining only the closest objects to cluster the far correlation between parameters can be restored thanks to effects of self-organization in databases. The application of the developed pattern recognition method to anonymized, consecutively collected archive data bases allows to restore the regulatory characteristics of complex systems in the form of the families of dependencies in attribute space. At the same time it does not need to use the special investigations of patients or biological objects and test influences on them. The other special conditions of data collecting are not necessary.