Early warning indicators via latent stochastic dynamical systems

Lingyu Feng,Ting Gao,Wang Xiao,Jinqiao Duan

DOI: https://doi.org/10.1063/5.0195042

2024-04-05

Abstract:Detecting early warning indicators for abrupt dynamical transitions in complex systems or high-dimensional observation data is essential in many real-world applications, such as brain diseases, natural disasters, and engineering reliability. To this end, we develop a novel approach: the directed anisotropic diffusion map that captures the latent evolutionary dynamics in the low-dimensional manifold. Then three effective warning signals (Onsager-Machlup Indicator, Sample Entropy Indicator, and Transition Probability Indicator) are derived through the latent coordinates and the latent stochastic dynamical systems. To validate our framework, we apply this methodology to authentic electroencephalogram (EEG) data. We find that our early warning indicators are capable of detecting the tipping point during state transition. This framework not only bridges the latent dynamics with real-world data but also shows the potential ability for automatic labeling on complex high-dimensional time series.

Machine Learning

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve

The paper aims to identify early warning signals of epileptic seizures through the following methods:

1. **Low-Dimensional Coordinate Search**: Finding appropriate low-dimensional coordinate representations from high-dimensional time series data.
2. **Implicit Stochastic Dynamical System Modeling**: Learning implicit stochastic differential equations and their dynamics on a low-dimensional manifold.
3. **Exploration of Effective Early Warning Indicators**: Developing three effective warning signals (Onsager-Machlup indicator, sample entropy indicator, and transition probability indicator) to detect state transitions in brain activity.

Specifically, the authors propose a new method—directed anisotropic diffusion map—to capture the underlying evolutionary dynamics on a low-dimensional manifold. They then derive three effective warning signals through latent coordinates and latent stochastic dynamical systems and apply these methods to real electroencephalogram (EEG) data to validate the framework's effectiveness. This framework not only links latent dynamics with real-world data but also demonstrates the potential to automatically label complex high-dimensional time series.

### Summary

The main goal of this study is to automatically identify early warning signals of epileptic seizures from brain activity data of epilepsy patients, using tools from stochastic dynamical systems to achieve this goal. Its main contributions include:
- Finding appropriate low-dimensional coordinates for high-dimensional time series data;
- Learning implicit stochastic differential equations and dynamics on a low-dimensional manifold;
- Exploring effective early warning indicators for brain activity transitions.