Inferring networks from time series: a neural approach

Thomas Gaskin,Grigorios A. Pavliotis,Mark Girolami

2023-11-01

Abstract:Network structures underlie the dynamics of many complex phenomena, from gene regulation and foodwebs to power grids and social media. Yet, as they often cannot be observed directly, their connectivities must be inferred from observations of the dynamics to which they give rise. In this work we present a powerful computational method to infer large network adjacency matrices from time series data using a neural network, in order to provide uncertainty quantification on the prediction in a manner that reflects both the degree to which the inference problem is underdetermined as well as the noise on the data. This is a feature that other approaches have hitherto been lacking. We demonstrate our method's capabilities by inferring line failure locations in the British power grid from its response to a power cut, providing probability densities on each edge and allowing the use of hypothesis testing to make meaningful probabilistic statements about the location of the cut. Our method is significantly more accurate than both Markov-chain Monte Carlo sampling and least squares regression on noisy data and when the problem is underdetermined, while naturally extending to the case of non-linear dynamics, which we demonstrate by learning an entire cost matrix for a non-linear model of economic activity in Greater London. Not having been specifically engineered for network inference, this method in fact represents a general parameter estimation scheme that is applicable to any high-dimensional parameter space.

Machine Learning,Optimization and Control

What problem does this paper attempt to address?

The problem this paper attempts to address is inferring the adjacency matrix of large networks from time series data and quantifying the uncertainty of the prediction results. Specifically:

- **Main Objective**: Propose a robust computational method that uses neural networks to infer the adjacency matrix of large networks from time series data and provide a quantification of prediction uncertainty, which can reflect the degree of underdetermination of the inference problem and data noise.

- **Application Scenarios**:
- Inferring the location of line faults in the UK power grid by observing the grid's response to power outage events to determine the fault location.
- Inferring the economic cost relationship network in the London area from economic activity data.

- **Technical Means**: Employ neural network methods combined with the idea of Bayesian inference, which can handle not only linear dynamical systems but also extend to parameter estimation problems of nonlinear dynamic systems.

- **Advantages**: Compared to traditional Markov Chain Monte Carlo (MCMC) sampling and least squares regression methods, this method performs better under noisy data and can still provide reliable uncertainty quantification results under underdetermined conditions.

In summary, this research aims to efficiently and accurately infer complex network structures from limited and noisy time series data and evaluate the reliability of the results.