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

Future surveys such as the Legacy Survey of Space and Time (LSST) of the VeraC. Rubin Observatory will observe an order of magnitude more astrophysicaltransient events than any previous survey before. With this deluge ofphotometric data, it will be impossible for all such events to be classified byhumans alone. Recent efforts have sought to leverage machine learning methodsto tackle the challenge of astronomical transient classification, with everimproving success. Transformers are a recently developed deep learningarchitecture, first proposed for natural language processing, that have shown agreat deal of recent success. In this work we develop a new transformerarchitecture, which uses multi-head self attention at its core, for generalmulti-variate time-series data. Furthermore, the proposed time-seriestransformer architecture supports the inclusion of an arbitrary number ofadditional features, while also offering interpretability. We apply thetime-series transformer to the task of photometric classification, minimisingthe reliance of expert domain knowledge for feature selection, while achievingresults comparable to state-of-the-art photometric classification methods. Weachieve a logarithmic-loss of 0.507 on imbalanced data in a representativesetting using data from the Photometric LSST Astronomical Time-SeriesClassification Challenge (PLAsTiCC). Moreover, we achieve a micro-averagedreceiver operating characteristic area under curve of 0.98 and micro-averagedprecision-recall area under curve of 0.87.

Paying Attention to Astronomical Transients: Introducing the Time-series Transformer for Photometric Classification