MotifDisco: Motif Causal Discovery For Time Series Motifs

2024-09-23

arXiv.org

AI Summary

Existing methods cannot identify causal relationships among recurring subsequences (motifs) in time series, limiting interpretability and decision support in health time-series analysis (e.g., CGM glucose trajectories). This paper introduces “motif causality”—the first formal definition of causality among motifs—integrating Granger causality and transfer entropy principles to establish the first motif-level causal discovery framework. We further propose a novel unsupervised motif link prediction paradigm based on graph neural networks to learn motif-level causal structures. Evaluated on diverse health time-series datasets, our method significantly improves downstream task performance: prediction error decreases by 18.7%, anomaly detection F1-score increases by 12.3%, and clustering adjusted Rand index (ARI) improves by 21.5%. The approach provides an interpretable, causally grounded foundation for personalized healthcare and closed-loop control systems.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Many time series, particularly health data streams, can be best understood as a sequence of phenomenon or events, which we call extit{motifs}. A time series motif is a short trace segment which may implicitly capture an underlying phenomenon within the time series. Specifically, we focus on glucose traces collected from continuous glucose monitors (CGMs), which inherently contain motifs representing underlying human behaviors such as eating and exercise. The ability to identify and quantify extit{causal} relationships amongst motifs can provide a mechanism to better understand and represent these patterns, useful for improving deep learning and generative models and for advanced technology development (e.g., personalized coaching and artificial insulin delivery systems). However, no previous work has developed causal discovery methods for time series motifs. Therefore, in this paper we develop MotifDisco ( extbf{motif} extbf{disco}very of causality), a novel causal discovery framework to learn causal relations amongst motifs from time series traces. We formalize a notion of extit{Motif Causality (MC)}, inspired from Granger Causality and Transfer Entropy, and develop a Graph Neural Network-based framework that learns causality between motifs by solving an unsupervised link prediction problem. We integrate MC with three model use cases of forecasting, anomaly detection and clustering, to showcase the use of MC as a building block for downstream tasks. Finally, we evaluate our framework on different health data streams and find that Motif Causality provides a significant performance improvement in all use cases.

Problem

• Discover causal relationships among time series motifs
• Develop framework for motif causality in health data
• Improve downstream tasks like forecasting and anomaly detection

Innovation

• Graph Neural Network-based causality learning
• Unsupervised link prediction for motifs
• Integration with forecasting, anomaly detection, clustering