Beyond Point Matching: Evaluating Multiscale Dubuc Distance for Time Series Similarity
This work addresses the challenge of efficient time series search and indexing in data mining, offering a novel similarity measure that is incremental but shows practical utility.
This paper evaluates the Multiscale Dubuc Distance (MDD) for time series similarity, comparing it to Dynamic Time Warping (DTW) and showing that MDD outperforms DTW in many scenarios with substantial gains, including a significant improvement in a real-world classification task.
Time series are high-dimensional and complex data objects, making their efficient search and indexing a longstanding challenge in data mining. Building on a recently introduced similarity measure, namely Multiscale Dubuc Distance (MDD), this paper investigates its comparative strengths and limitations relative to the widely used Dynamic Time Warping (DTW). MDD is novel in two key ways: it evaluates time series similarity across multiple temporal scales and avoids point-to-point alignment. We demonstrate that in many scenarios where MDD outperforms DTW, the gains are substantial, and we provide a detailed analysis of the specific performance gaps it addresses. We provide simulations, in addition to the 95 datasets from the UCR archive, to test our hypotheses. Finally, we apply both methods to a challenging real-world classification task and show that MDD yields a significant improvement over DTW, underscoring its practical utility.