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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The detection of anomalies in multivariate time series data is crucial forvarious practical applications, including smart power grids, traffic flowforecasting, and industrial process control. However, real-world time seriesdata is usually not well-structured, posting significant challenges to existingapproaches: (1) The existence of missing values in multivariate time seriesdata along variable and time dimensions hinders the effective modeling ofinterwoven spatial and temporal dependencies, resulting in important patternsbeing overlooked during model training; (2) Anomaly scoring withirregularly-sampled observations is less explored, making it difficult to useexisting detectors for multivariate series without fully-observed values. Inthis work, we introduce a novel framework called GST-Pro, which utilizes agraph spatiotemporal process and anomaly scorer to tackle the aforementionedchallenges in detecting anomalies on irregularly-sampled multivariate timeseries. Our approach comprises two main components. First, we propose a graphspatiotemporal process based on neural controlled differential equations. Thisprocess enables effective modeling of multivariate time series from bothspatial and temporal perspectives, even when the data contains missing values.Second, we present a novel distribution-based anomaly scoring mechanism thatalleviates the reliance on complete uniform observations. By analyzing thepredictions of the graph spatiotemporal process, our approach allows anomaliesto be easily detected. Our experimental results show that the GST-Pro methodcan effectively detect anomalies in time series data and outperformsstate-of-the-art methods, regardless of whether there are missing valuespresent in the data. Our code is available: https://github.com/huankoh/GST-Pro.

Graph Spatiotemporal Process for Multivariate Time Series Anomaly Detection with Missing Values