Unsupervised learning for anticipating critical transitions
This work addresses a key limitation in predicting critical transitions for complex systems, offering an unsupervised method that could benefit fields like climate science or engineering, though it appears incremental as it builds on existing reservoir computing approaches.
The paper tackles the problem of anticipating critical transitions in complex dynamical systems without requiring explicit knowledge of the bifurcation parameter, by combining a variational autoencoder (VAE) with reservoir computing to detect driving factors from time series and achieve accurate predictions, as demonstrated on systems like the Kuramoto-Sivashinsky model.
For anticipating critical transitions in complex dynamical systems, the recent approach of parameter-driven reservoir computing requires explicit knowledge of the bifurcation parameter. We articulate a framework combining a variational autoencoder (VAE) and reservoir computing to address this challenge. In particular, the driving factor is detected from time series using the VAE in an unsupervised-learning fashion and the extracted information is then used as the parameter input to the reservoir computer for anticipating the critical transition. We demonstrate the power of the unsupervised learning scheme using prototypical dynamical systems including the spatiotemporal Kuramoto-Sivashinsky system. The scheme can also be extended to scenarios where the target system is driven by several independent parameters or with partial state observations.