Abstract:The similarity search on sensor data generated by a myriad of sensing devices is a frequently encountered problem in the era of the Internet of Things (IoT). This sensor data generally appear in the form of time series, a temporally ordered sequence of real numbers obtained regularly in time. It has been widely accepted that the dynamic time warping (DTW) currently is the most prevalent similarity measure in the time-series mining community, mainly due to its flexibility and broad applicability. However, calculating DTW between two time series has quadratic time complexity, leading to unsatisfactory efficiency when performing the similarity search over the large time-series data set. The main contribution of this article is to propose a method called product quantization (PQ)-based DTW (PQDTW) for fast time-series approximate similarity search under DTW. The PQ, a well-known approximate nearest neighbor search approach, is used in PQDTW. Nevertheless, the conventional PQ is developed with the Euclidean distance and is not designed for DTW. To solve this problem, the DTW barycenter averaging (DBA) technique is utilized to adapt the PQ for DTW before using it. We employ PQDTW along with the filter-and-refine framework to efficiently and accurately perform the time-series similarity search. Our method can reasonably reduce many DTW computations in the filtering phase; thus, the query process is accelerated. We compare PQDTW with related popular algorithms using public time-series data sets. Experimental results verify that the proposal achieves the best tradeoff between query efficiency and retrieval accuracy compared to the competitors.

Piecewise Chebyshev Factorization Based Nearest Neighbour Classification for Time Series

Piecewise Factorization for Time Series Classification.

Multilevel Dynamic Time Warping: A Parameter-Light Method for Fast Time Series Classification

A novel piecewise linear segmentation for time series

Dynamic Time Warping under Product Quantization, with Applications to Time-Series Data Similarity Search

An efficient method for time series similarity search using binary code representation and hamming distance

Integrating Data-Driven Segmentation, Local Feature Extraction and Fisher Kernel Encoding to Improve Time Series Classification

Piecewise Statistic Approximation Based Similarity Measure for Time Series

A Novel Similarity Measure Approach for Time Series Based on PLA and DTW

An Experimental Evaluation of Nearest Neighbour Time Series Classification

A Novel Similarity Measure Model for Multivariate Time Series Based on LMNN and DTW

Nearest Neighbor Optimal Smooth Denoising Dynamic Classification Method for Financial Time Series

Using dynamic time warping distances as features for improved time series classification

Convolutional Nonlinear Neighbourhood Components Analysis for Time Series Classification.

Exploiting Multi-Channels Deep Convolutional Neural Networks for Multivariate Time Series Classification

Dynamic Time Warping: Itakura vs Sakoe-Chiba

DTW-NN: A novel neural network for time series recognition using dynamic alignment between inputs and weights

Supporting Subseries Nearest Neighbor Search Via Approximation

Affine and Regional Dynamic Time Warpng

An Interpretable Time Series Classification Approach Based on Feature Clustering.

Entropy-Based Symbolic Representation for Time Series Classification