$k$-Graph: A Graph Embedding for Interpretable Time Series Clustering
This addresses the need for more interpretable clustering in time series analysis across various domains, though it appears incremental as it builds on existing graph-based approaches.
The paper tackles the problem of limited interpretability in time series clustering by introducing $k$-Graph, an unsupervised method that uses graph representations of subsequences to improve interpretability and accuracy, outperforming state-of-the-art algorithms.
Time series clustering poses a significant challenge with diverse applications across domains. A prominent drawback of existing solutions lies in their limited interpretability, often confined to presenting users with centroids. In addressing this gap, our work presents $k$-Graph, an unsupervised method explicitly crafted to augment interpretability in time series clustering. Leveraging a graph representation of time series subsequences, $k$-Graph constructs multiple graph representations based on different subsequence lengths. This feature accommodates variable-length time series without requiring users to predetermine subsequence lengths. Our experimental results reveal that $k$-Graph outperforms current state-of-the-art time series clustering algorithms in accuracy, while providing users with meaningful explanations and interpretations of the clustering outcomes.