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

Multivariate long sequence time-series forecasting (M-LSTF) is a practicalbut challenging problem. Unlike traditional timer-series forecasting tasks,M-LSTF tasks are more challenging from two aspects: 1) M-LSTF models need tolearn time-series patterns both within and between multiple time features; 2)Under the rolling forecasting setting, the similarity between two consecutivetraining samples increases with the increasing prediction length, which makesmodels more prone to overfitting. In this paper, we propose a generalizablememory-driven Transformer to target M-LSTF problems. Specifically, we firstpropose a global-level memory component to drive the forecasting procedure byintegrating multiple time-series features. In addition, we adopt a progressivefashion to train our model to increase its generalizability, in which wegradually introduce Bernoulli noises to training samples. Extensive experimentshave been performed on five different datasets across multiple fields.Experimental results demonstrate that our approach can be seamlessly pluggedinto varying Transformer-based models to improve their performances up toroughly 30%. Particularly, this is the first work to specifically focus on theM-LSTF tasks to the best of our knowledge.

Generalizable Memory-driven Transformer for Multivariate Long Sequence Time-series Forecasting