Unsupervised learning for anticipating critical transitions

TL;DR

This paper introduces an unsupervised learning framework combining variational autoencoders and reservoir computing to predict critical transitions in complex dynamical systems without prior knowledge of bifurcation parameters.

Contribution

The novel framework detects driving factors from time series using VAE and uses this information for transition prediction, extending prior methods that required explicit parameter knowledge.

Findings

Successfully predicts critical transitions in prototypical systems

Works with multiple parameters and partial observations

Demonstrates effectiveness on the Kuramoto-Sivashinsky system

Abstract

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.