Spontaneous data clustering using collective synchronization in a network of phase oscillators

We developed a method for spontaneous data clustering based on Kuramoto’s model for collective synchronization. A network of phase oscillators, to the natural frequencies of which multivariate data are input, achieves partial synchrony owing to short range interaction between neighboring phase oscillators. The common frequencies of the partial synchronous groups represent major feature patterns of the multivariate data. As a case study, we apply our method to actually observed time series of wind velocity and show major feature patterns of the wind data.