Towards Explainable Deep Clustering for Time Series Data

TL;DR

The paper addresses the need for explainable deep clustering in time series by surveying current methods and organizing them along intrinsic and post-hoc explainability strategies. It highlights the dominance of autoencoder-plus-attention architectures and identifies gaps in streaming, privacy-preserving, and domain-tailored explanations across healthcare, finance, IoT, and climate science. By detailing attention-based, prototype/exemplar, SOM-based, and post-hoc visualization approaches, it outlines six concrete research opportunities to elevate interpretability from an afterthought to a core design goal. The work aims to guide the development of trustworthy, practically deployable explainable clustering systems that can support critical decision-making in real-world settings.

Abstract

Deep clustering uncovers hidden patterns and groups in complex time series data, yet its opaque decision-making limits use in safety-critical settings. This survey offers a structured overview of explainable deep clustering for time series, collecting current methods and their real-world applications. We thoroughly discuss and compare peer-reviewed and preprint papers through application domains across healthcare, finance, IoT, and climate science. Our analysis reveals that most work relies on autoencoder and attention architectures, with limited support for streaming, irregularly sampled, or privacy-preserved series, and interpretability is still primarily treated as an add-on. To push the field forward, we outline six research opportunities: (1) combining complex networks with built-in interpretability; (2) setting up clear, faithfulness-focused evaluation metrics for unsupervised explanations; (3) building explainers that adapt to live data streams; (4) crafting explanations tailored to specific domains; (5) adding human-in-the-loop methods that refine clusters and explanations together; and (6) improving our understanding of how time series clustering models work internally. By making interpretability a primary design goal rather than an afterthought, we propose the groundwork for the next generation of trustworthy deep clustering time series analytics.