Early Abandoning PrunedDTW and its application to similarity search

TL;DR

This paper introduces EAPrunedDTW, an improved version of PrunedDTW, which significantly accelerates DTW-based similarity search by integrating early abandoning from the outset, eliminating the need for lower bounds.

Contribution

The paper presents EAPrunedDTW, a novel algorithm that enhances PrunedDTW with built-in early abandoning, leading to faster DTW computations for time series similarity search.

Findings

EAPrunedDTW outperforms previous methods in speed.

It reduces reliance on lower bounds in DTW.

Significant improvements in UCR Suite search times.

Abstract

The Dynamic Time Warping ("DTW") distance is widely used in time series analysis, be it for classification, clustering or similarity search. However, its quadratic time complexity prevents it from scaling. Strategies, based on early abandoning DTW or skipping its computation altogether thanks to lower bounds, have been developed for certain use cases such as nearest neighbour search. But vectorization and approximation aside, no advance was made on DTW itself until recently with the introduction of PrunedDTW. This algorithm, able to prune unpromising alignments, was later fitted with early abandoning. We present a new version of PrunedDTW, "EAPrunedDTW", designed with early abandon in mind from the start, and able to early abandon faster than before. We show that EAPrunedDTW significantly improves the computation time of similarity search in the UCR Suite, and renders lower bounds dispensable.