FastDTW is Approximate and Generally Slower Than the Algorithm it Approximates

Many time series data mining problems can be solved with repeated use of distance measure. Examples such tasks include similarity search, clustering, classification, anomaly detection and segmentation. For over two decades it has been known that the Dynamic Time Warping (DTW) measure is best to for most tasks, in domains. Because classic DTW algorithm quadratic complexity, many ideas have introduced reduce its amortized time, or quickly approximate it. One cited approaches FastDTW. The FastDTW well a thousand citations explicitly used several hundred research efforts. In this work, we make surprising claim. any realistic application, approximate much slower than xmlns:xlink="http://www.w3.org/1999/xlink">exact DTW. This fact clearly implications community uses algorithm: allowing address larger datasets, get exact results, do so less time.