English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Suppose we are given a system of coupled oscillators on an unknown graphalong with the trajectory of the system during some period. Can we predictwhether the system will eventually synchronize? Even with a known underlyinggraph structure, this is an important yet analytically intractable question ingeneral. In this work, we take an alternative approach to the synchronizationprediction problem by viewing it as a classification problem based on the factthat any given system will eventually synchronize or converge to anon-synchronizing limit cycle. By only using some basic statistics of theunderlying graphs such as edge density and diameter, our method can achieveperfect accuracy when there is a significant difference in the topology of theunderlying graphs between the synchronizing and the non-synchronizing examples.However, in the problem setting where these graph statistics cannot distinguishthe two classes very well (e.g., when the graphs are generated from the samerandom graph model), we find that pairing a few iterations of the initialdynamics along with the graph statistics as the input to our classificationalgorithms can lead to significant improvement in accuracy; far exceeding whatis known by the classical oscillator theory. More surprisingly, we find that inalmost all such settings, dropping out the basic graph statistics and trainingour algorithms with only initial dynamics achieves nearly the same accuracy. Wedemonstrate our method on three models of continuous and discrete coupledoscillators -- the Kuramoto model, Firefly Cellular Automata, andGreenberg-Hastings model. Finally, we also propose an "ensemble prediction"algorithm that successfully scales our method to large graphs by training ondynamics observed from multiple random subgraphs.

Learning to predict synchronization of coupled oscillators on randomly generated graphs