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

Time series imputation remains a significant challenge across many fields dueto the potentially significant variability in the type of data being modelled.Whilst traditional imputation methods often impose strong assumptions on theunderlying data generation process, limiting their applicability, researchershave recently begun to investigate the potential of deep learning for thistask, inspired by the strong performance shown by these models in bothclassification and regression problems across a range of applications. In thiswork we propose MADS, a novel auto-decoding framework for time seriesimputation, built upon implicit neural representations. Our method leveragesthe capabilities of SIRENs for high fidelity reconstruction of signals andirregular data, and combines it with a hypernetwork architecture which allowsus to generalise by learning a prior over the space of time series. We evaluateour model on two real-world datasets, and show that it outperformsstate-of-the-art methods for time series imputation. On the human activitydataset, it improves imputation performance by at least 40%, while on the airquality dataset it is shown to be competitive across all metrics. Whenevaluated on synthetic data, our model results in the best average rank acrossdifferent dataset configurations over all baselines.

MADS: Modulated Auto-Decoding SIREN for time series imputation