Inferring the time-varying coupling of dynamical systems with temporal convolutional autoencoders

TL;DR

This paper introduces TACI, a novel machine learning approach using temporal autoencoders to accurately infer and measure dynamic causal interactions in complex, non-linear, and non-stationary systems, outperforming existing methods.

Contribution

The paper presents TACI, a new methodology combining a surrogate data metric with a two-headed neural network architecture for causal inference in time-varying systems.

Findings

TACI accurately quantifies dynamic causal interactions in synthetic datasets.

TACI outperforms existing methods on real-world data.

The approach reveals insights into mechanisms of time-varying interactions.

Abstract

Most approaches for assessing causality in complex dynamical systems fail when the interactions between variables are inherently non-linear and non-stationary. Here we introduce Temporal Autoencoders for Causal Inference (TACI), a methodology that combines a new surrogate data metric for assessing causal interactions with a novel two-headed machine learning architecture to identify and measure the direction and strength of time-varying causal interactions. Through tests on both synthetic and real-world datasets, we demonstrate TACI's ability to accurately quantify dynamic causal interactions across a variety of systems. Our findings display the method's effectiveness compared to existing approaches and also highlight our approach's potential to build a deeper understanding of the mechanisms that underlie time-varying interactions in physical and biological systems.