Time series causal discovery methods explained and unified across fields
One Thing Leads to Another: The Essentials of Time Series Causal Discovery
Machine Learning
Summary
Finding cause-and-effect links in data that changes over time is a complex and growing challenge. The authors provide a clear and organized overview of how different fields like signal processing, statistics, and machine learning approach this problem. They explain key ideas, assumptions, and the limits of current methods. They also talk about how these methods are tested and used in areas like farming, healthcare, and AI development.
What this means in practice
- •For agriculture teams: Use causal discovery methods to identify factors affecting crop yields over time for better farming decisions.
- •For healthcare data analysts: Detect causal relationships between patient vital signs and treatments to improve clinical decision-making.
A survey. It maps existing work.
Authors
Kurt Butler, Damian Machlanski, Michael P. J. Camilleri, Zhihua Liu, Panagiotis Dimitrakopoulos, Feng Chen, Ricardo Silva, Sotirios A. Tsaftaris
Abstract
Time series causal discovery --- the process of detecting causation between time series --- has its roots in signal processing research from over a half century ago. It has since received active and widespread interest, evolving into different forms across domains such as statistics, dynamical systems and machine learning. Entering this fragmented and constantly evolving research landscape can be daunting. This tutorial provides a unified overview of time series causal discovery across research communities. We introduce the key principles invoked by modern discovery methods, and we discuss their limitations and assumptions. We also discuss how discovery methods are evaluated, benchmarked and fine-tuned. We further discuss how causal discovery for time series is becoming crucial for downstream tasks in agriculture and Earth science, in healthcare, and in artificial intelligence systems. We believe that this tutorial will serve as a valuable resource in navigating the deluge of recent causality research, as well as an invitation to signal processing audiences to contribute to causal problems. To further help readers, we host a companion repository to link to useful code packages and resources at https://github.com/CHAI-UK/Awesome-Time-Series-Causal-Discovery.