English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Time series classification is an important data mining task that has receiveda lot of interest in the past two decades. Due to the label scarcity inpractice, semi-supervised time series classification with only a few labeledsamples has become popular. Recently, Similarity-aware Time SeriesClassification (SimTSC) is proposed to address this problem by using a graphneural network classification model on the graph generated from pairwiseDynamic Time Warping (DTW) distance of batch data. It shows excellent accuracyand outperforms state-of-the-art deep learning models in several few-labelsettings. However, since SimTSC relies on pairwise DTW distances, the quadraticcomplexity of DTW limits its usability to only reasonably sized datasets. Toaddress this challenge, we propose a new efficient semi-supervised time seriesclassification technique, LB-SimTSC, with a new graph construction module.Instead of using DTW, we propose to utilize a lower bound of DTW, LB_Keogh, toapproximate the dissimilarity between instances in linear time, while retainingthe relative proximity relationships one would have obtained via computing DTW.We construct the pairwise distance matrix using LB_Keogh and build a graph forthe graph neural network. We apply this approach to the ten largest datasetsfrom the well-known UCR time series classification archive. The resultsdemonstrate that this approach can be up to 104x faster than SimTSC whenconstructing the graph on large datasets without significantly decreasingclassification accuracy.

LB-SimTSC: An Efficient Similarity-Aware Graph Neural Network for Semi-Supervised Time Series Classification