Extracting Feature Patterns in the Health Status of Elderly People Needing Nursing Care by Data Synchronization

Abstract

We devised a method for data mining from multivariate data using a network of coupled phase oscillators subject to an analogue of the Kuramoto model for collective synchronization. In our method, the natural frequencies of the phase oscillators are extended to vector quantities to which multivariate data are assigned. The common frequency vectors of partially synchronized groups of phase oscillators are interpreted to be the template vectors representing the major features of the data set. We applied our method to care-needs-certification data in the Japanese public long-term care insurance program, and extracted major patterns in the health status of the elderly needing nursing care and their dependence on the model parameter representing the level of coarse-graining for data clustering.