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 Transformer is a highly successful deep learning model that hasrevolutionised the world of artificial neural networks, first in naturallanguage processing and later in computer vision. This model is based on theattention mechanism and is able to capture complex semantic relationshipsbetween a variety of patterns present in the input data. Precisely because ofthese characteristics, the Transformer has recently been exploited for timeseries forecasting problems, assuming a natural adaptability to the domain ofcontinuous numerical series. Despite the acclaimed results in the literature,some works have raised doubts about the robustness and effectiveness of thisapproach. In this paper, we further investigate the effectiveness ofTransformer-based models applied to the domain of time series forecasting,demonstrate their limitations, and propose a set of alternative models that arebetter performing and significantly less complex. In particular, we empiricallyshow how simplifying Transformer-based forecasting models almost always leadsto an improvement, reaching state of the art performance. We also proposeshallow models without the attention mechanism, which compete with the overallstate of the art in long time series forecasting, and demonstrate their abilityto accurately predict time series over extremely long windows. From amethodological perspective, we show how it is always necessary to use a simplebaseline to verify the effectiveness of proposed models, and finally, weconclude the paper with a reflection on recent research paths and theopportunity to follow trends and hypes even where it may not be necessary.

Two Steps Forward and One Behind: Rethinking Time Series Forecasting with Deep Learning