On using the modularity of recurrence network communities to detect change-point behaviour

A new network algorithm to detect when time series changes concepts.

Recurrence network communities as a proxy for quadrant scan tipping point detection.

Real-world industrial application to streaming data and run-to-failure experiments.